Prediction of the Fatty Acid Profiles of Iberian Pig Products by Near Infrared Spectroscopy: A Comparison Between Multiple Regression Tools and Artificial Neural Networks

Prediction of the Fatty Acid Profiles of Iberian Pig Products by Near Infrared Spectroscopy: A Comparison Between Multiple Regression Tools and Artificial Neural Networks | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Prediction of the Fatty Acid Profiles of Iberian Pig Products by Near Infrared Spectroscopy: A Comparison Between Multiple Regression Tools and Artificial Neural Networks Miriam Hernández-Jiménez, Isabel Revilla, Pedro Hernández-Ramos, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4241621/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract In this study the feasibility of predicting the lipid profiles of Iberian ham and shoulder samples by using NIRS technology was evaluated. Gas chromatography analysis was the reference method used. The muscles analyzed and recorded by NIRS were 76 Biceps femoris for Iberian hams and 72 Brachiocephalicus for Iberian shoulders. NIR calibrations were carried out by using two methods: modified partial least squares regression (MPLS) and artificial neural networks (ANN). With the MPLS method it was possible to obtain equations with RSQ of > 0.5 for 5 individual fatty acids and 3 summations (PUFA, n-3 and n-6). The use of neural networks made it possible to find equations with RSQ of > 0.5 for 10 individual fatty acids, all of which are present in over 90% of the samples, and 5 summates (SFA, MUFA, PUFA, n-3 and n-6); finding that the calibration curves of the fatty acids C18:1, C18:2n6 and C18:3n3 presented RSQs of > 0.7. The results obtained indicate that NIR spectroscopy could be a very useful technology for the quality control of cured products as it allows estimating the main fatty constituents quickly and without using reagents. Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Iberian ham is one of Spain's most renowned products (Toldrá & Aristoy, 2004 ). Ham and all other products made from the Iberian pig are characterized by their high percentage of fat, which is a key factor in its influence on technological quality and its impact on the sensory quality of the final products (Gandemer, 2002 ). This is the reason why so much research has been carried out to define the production factors affecting the intramuscular fat content and the composition of this fat. These factors include the breed, age, and live weight of the animal, but diet composition in the pre-slaughter period is considered to be the most influential (Carrapiso, Bonilla & García, 2003). The higher quality of cured products from the Iberian breed is usually correlated with the traditional production systems which consist of feeding these Iberian pigs during the fattening period ( montanera ) exclusively on natural resources found in a traditional landscape known as the dehesa , i.e. pastures and mainly acorns (Rodríguez-Estévez et al., 2009 ). Current Spanish legislation classifies Iberian pig products (ham, shoulder, and loin) according to four categories which are Bellota 100% for products of the highest quality followed by Bellota Ibérico, Ibérico de cebo de campo, and Ibérico de cebo (RD 4/2014). The first two categories refer to animals fed during the montanera (grazing free-range on grass and acorns) but the 100% Ibérico come from purebred Iberian animals and the Bellota Ibérico from Iberian x Duroc crossbred animals. This system of montanera allows the pork to acquire higher intramuscular fat with a characteristic high proportion of unsaturated fatty acids (Cava et al., 1997 ; González & Tejeda, 2007 ; Ruiz-Carrascal, Ventanas, Cava, Andrés & García, 2000; Tejeda, Gandemer, Antequera, Viau, & García 2002) which contributes to the unique and desirable sensory properties of the product. However, the higher degree of fat unsaturation also increases lipid oxidation and thus decreases shelf life (Rubio et al., 2007 ; Shahidi & Zhong, 2010 ; Sheard et al., 2000 ). On the one hand, it is nowadays strongly recommendable to reduce the intake of saturated fats and increase the intake of polyunsaturated fats, especially omega-3 fats, to prevent cardiovascular disease (Mataix, Quiles & Rodríguez, 2001 ). Consumers have therefore shown an increasing interest in the fat composition of their diets and demand more extensive knowledge of the fatty acid profiles of the products they purchase. On the other hand, the Iberian pork industry is seeking quality control systems which allow it to certify the authenticity of its products at any point in the distribution chain and provide accurate product information, setting the price according to quality (Fernández-Cabanás et al., 2011 ). Ascertaining the fatty acid profiles of products such as ham and shoulder could provide important information related to quality control, sensory attributes, shelf life, and nutritional properties (Fernández-Cabanás et al., 2011 ; Tejerina, García-Torres, Cabeza de Vaca, Ortiz & Romero-Fernández, 2018). The analysis of quality parameters in general and the lipid profile in particular involves sample destruction and is expensive and time-consuming, which is a problem for the meat industry (Tejerina et al., 2018 ). Non-invasive technologies are therefore useful alternatives for quality analysis and product authentication. In this scenario, low-field 1H-NMR has been used to determine the lipid profile of Iberian ham; furthermore, classification according to current commercial categories was possible using chemometric tools (Pajuelo et al., 2022 ). Ion Mobility Spectrometry coupled with Gas Chromatography (IMS-GC) and the use of non-destructive sampling methods has also been shown to be capable of classifying ham samples according to their commercial category by using volatilome (Martín-Gómez, Arroyo-Manzanares, Rodríguez-Estévez & Arce, 2019 ). Near Infrared Spectroscopy (NIRS) is one of the most outstanding non-invasive technologies as it is a fast, non-destructive, and accurate tool for predicting the chemical composition of meat (Alomar, Gallo, Castañeda & Fuchslocher, 2003 ). The main advantage of NIR spectroscopy is that it allows a multitude of analytical parameters to be determined in a few minutes, facilitating a more exhaustive control of a large number of individuals or pieces (Prieto, Pawludzyk, Dugan& Aalhus, 2017; Zamora-Rojas, Garrido-Varo, De Pedro-Sanz, Guerrero-Ginel & Pérez-Marín, 2011 ). In the Iberian sector several research studies have been carried out using quantitative models which have shown a high potential for predicting the fatty acid and stable carbon isotope profile in subcutaneous fat of the carcass and/or live animal (García-Olmo, Garrido-Varo, De Pedro, 2001 ; González-Martín et al., 2021 ; Pérez-Marín et al., 2009 ). Qualitative models have also allowed classification according to the diet, breed and management systems by using spectra recorded both in the live animal and in different areas of the carcass (Arce et al., 2009 ; Hernández-Jiménez, González-Martín, Martínez-Martín, Revilla & Vivar-Quintana 2021 ; Horcada, Valera, Juárez & Fernández-Cabanás, 2020 ; Pérez-Marín, Fearn, Riccioli, De Pedro & Garrido, 2021 ). The use of NIR spectroscopy in cured products has allowed the prediction of the main quality characteristics in a fast and non-destructive manner (Ortiz, León, Contador & Tejerina, 2021 ). However, studies of dry-cured ham are rather scarce and include the prediction of texture and color (García, Andrada, Muñoz & Bote, 2005 ), the sensory characteristics (Hernández-Ramos et al., 2020 ) or salt content, and the fatty acid oxidation and tocopherol content in sliced and packaged ham (Tejerina et al., 2018 ) while the investigation of the feasibility of fatty acid profile prediction has been addressed by only a few studies (Fernández-Cabanás et al., 2011 ; Tejerina et al., 2018 ). The determination of the fatty acid profile of the final product by NIRS would facilitate the characterization of the product in the slicing and packaging stage. Furthermore, it would allow its authentication by relating the spectra to the diet and finally it would be a very useful tool for determining the fat saturation index so as to correlate it with the shelf life of the product (Ansorena & Astiasarán, 2004 ; Rubio et al., 2007 ; Sheard et al., 2000 ). One of the drawbacks of NIRS spectroscopy is data processing (Jurinjak Tušek et al., 2022 ). Several traditional methods are used such as Principal Component Analysis (PCA), Principal Component Regression (PCR), Canonical Correlation Analysis (CCA), Multiple Linear Regression (MLR), and in particular Partial Least Square Regression (PLSR) (Balabin & Safieva, 2011 ; Pérez-Marín, Garrido-Varo, De Pedro & Guerrero-Ginel, 2007 ). All of these are based on linear fit regressions. The nonlinearity of the data deriving from NIR spectroscopy is the main problem of these multivariate methods (Balabin, Safieva & Lomakina, 2007 , Balabin, Safieva & Lomakina, 2008 ). Therefore, the search for more complex algorithms with nonlinear and nonparametric regression has been resorted to; these algorithms include artificial neural networks (ANN) which have shown great potential (Pérez-Marín et al., 2007 ). ANNs are able to recognize and reproduce cause-effect relationships through a multitude of training in input-output systems (Kundu, Paul, Kumar & Mishra, 2015 ). In the agri-food sector they have been used in combination with NIRS spectroscopy for the prediction of sensory parameters in raw-cured meat products (Hernández-Jiménez et al., 2020 ; Hernández-Ramos et al., 2020 ), the prediction of textural parameters in cheese (Vásquez et al., 2018 ), and quality determination in fruits (Alhamdan, Fickak & Atia, 2019 ). No previous studies have been carried out with the aim of predicting fatty acids with neural networks in complex matrices such as meat, but studies of lipid profile determination in oils do exist (Rajković et al., 2023 ). The aim of this study was to determine the potential of NIR spectroscopy for the characterization of the lipid profile of Iberian ham and shoulder samples at different curing times prior to their commercialization. Two different methods were evaluated for the development of predictive models, one based on linear regression (MPLS) and the other based on non-linear regression (ANN). Several training algorithms and network architectures were evaluated for this latter method prior to the development of the model. 2. Materials and methods 2.1. Samples A total of 148 pieces (76 shoulders and 72 hams) from Iberian pigs slaughtered at 145 ± 15 kg carcass weight were selected for analysis. All the animals were fattened during the montanera , i.e. fed on grass and acorns of Quercus ilex and Quercus suber for 68–120 days depending on the management system of the farm of origin. Half of the animals were of 100% Iberian breed and the other half of 50% Iberian x Duroc breed. The pieces were processed by the local industry (Carrasco Ibéricos, Guijuelo, Spain) according to traditional methods consisting of a salting stage in salt piles at low temperature (1–4ºC) and high relative humidity (83–85%) for 1 day per kg of weight of the piece. At the end of this stage the surface was brushed and washed to remove the surface salt; the pieces subsequently underwent the post-salting stage for 9 months at temperatures of 4–5ºC and a relative humidity of 80%. They were then moved to drying rooms where the temperature was progressively increased from 4 to 27ºC while the relative humidity was decreased from 70–50% for 4–5 months. Finally, the hams and shoulders underwent a cellar stage with temperatures ranging from 17 to 25ºC and a relative humidity of 55–60%. Throughout this process intermediate sampling was carried out on a total of 16 shoulders at 250 days (8 months) and a total of 12 hams at 585 days (20 months). In addition, 60 pieces of shoulder and 60 pieces of ham were sampled at the end of the cellar process with processing times of 768 days (25 months) and 1075 days (36 months) respectively. For the sampling of the pieces, the bone was removed and a slice 3 cm thick was cut in the central area of the piece along its entire transverse profile, including all the muscles. The cut was made perpendicular to the bone and above it at the same height in all the pieces. This type of sampling allows the analysis of physical and chemical properties in addition to the NIRS recording of the different muscles identified. To analyze the lipid profile, the Biceps femoris muscles in the case of the hams and the Brachiocephalic muscle in the case of the shoulders were trimmed and kept at temperatures of -32ºC until the time of analysis. 2.2. Fatty acid analysis The fatty acid profile was determined by using gas chromatography. Total lipid extraction was carried out by using the method described by Folch, Lees & Sloane Stanley ( 1957 ), which combines mechanical extraction with organic solvent extraction in two different muscles, Biceps femoris in the case of hams and Brachiocephalicus in the case of shoulders. Subsequent methylation was performed on 0.1 grams of extracted fat by using the method described by González-Martín, Vicente Palacios, Revilla, Vivar-Quintana & Hernández-Hierro (2017). A GC 6890 N (Agilent Technologies, USA) chromatograph equipped with an automatic injector 7683 (Agilent Technologies) and a fused silica capillary column (100 m × 0.25 mm; 0.20 µm silica (SP-2560, Supelco, Inc., Bellefonte, PA, USA) were used. The injector and detector were maintained at 250ºC. The column oven temperature was 150ºC and the temperature was increased 1ºC/min to 165ºC, then 0.20ºC/min to 167ºC, and finally increased 1.50ºC/min to 225ºC for 15 min. The carrier gas was helium at a flow rate of 1ml/min. The injection volume was 1 µL in splitless mode. The different fatty acids were identified by their retention times using a mixture of fatty acid standards (47885-U Supelco, Sigma-Aldrich, Germany). Fatty acid contents were calculated by using the peak areas of the chromatograms and expressed as g per 100 g of the total fatty acid methyl esters. 2.3. Near infrared Spectroscopy (NIRS) spectra register The NIRS spectra of the samples obtained using a Foss NIRSystem 5000 (Hillerod, Denmark) were recorded using a fiber-optic probe (1.5 m 210/210, Ref. n° R6539-A) coupled with a 5 cm x 5 cm window quartz. The spectra of the sample were recorded in the 1100–2000 nm range at intervals (2 nm), which means that a total of 451 data were obtained for each sample and 32 scans were performed for each recording. The window was applied directly to the surface of the ham or shoulder slice. In the ham, the recording was performed on the Biceps femoris . In the shoulder, the spectral recording was performed on the Brachiocephalicus muscle. NIRS measurements were taken in diffuse reflectance and as a first step the spectra were converted from reflectance to absorbance as log 1/R. The three absorbance spectra of each sample were visually examined for consistency and then averaged (the outlier spectra were removed). The Win ISI 4.10 software package was used for spectra collection and data handling. 2.4. Calibration To predict the lipid profile jointly in both types of muscles ( Biceps femoris and Brachiocephalicus ), calibration was carried out by using two different methods in order to compare them. The first was the modified partial least squares regression method (MPLS) and the second was artificial neural networks (ANNs). Of the total fatty acids determined in ham and shoulder muscles at different curing times, we selected those that could be quantified in at least 90% of the samples analyzed. 2.4.1 MPLS Calibration The initial data for the calibrations were the 451 spectral data recorded in the near infrared region (1100–2000, 2 nm) together with the quantified percentage of each fatty acid (C14:0, C16:0, C18:0, C16:1, C18:1, C18:1 n7, C22:1 n9, C18:2 n6, C18:3 n3, and C20:2 n6) and their corresponding sums (SFA, MUFA, PUFA, n3 and n6) from each of the ham and shoulder samples (148 samples). These data sets were used as input and output data respectively. Possible anomalous or outlier spectra was detected by using principal component analysis (PCA) which also reduced the number of variables. The Mahalanobis distance indicates how different a spectrum is from the average spectrum of the set (De Maesschalck, Jouan-Rimbaud & Massart, 2000 ). When H took values of H > 3.5 the spectrum was considered anomalous and eliminated from the population. The criterion T > 2.5 was also applied in order to eliminate from the calibration set those samples with differences between the laboratory-determined fatty acid percentage and the NIRS predicted value. Because NIR spectra are affected by the physical characteristics of the sample, i.e. light scattering and wavelength variations, scattering treatments were applied including multiplicative scatter correction (MSC), standard normal variable (SNV), DeTrend (DT) or SNV-DT, and both first and second order derivatives (Marini, 2013 ; Norris, 2001 ). The regression method to obtain the lipid profile prediction equations using the NIR spectra as independent variables was the MPLS (modified partial least squares) method. MPLS is an improved form of PLS developed by FOSS and obtains better results in handling small spectral variability (Xie et al., 2022 ). The total spectra were divided into two subsets: a calibration set and a validation set. During the development of the different equations, a cross-validation process was carried out in order to select the optimal number of factors and avoid overfitting (Pérez-Marín et al., 2007 ) and the equation with the best predictive capacity was selected according to the following criteria: the highest determination coefficient in the calibration (RSQ or R2) and the lowest standard error calibration (SEC) and standard error cross-validation (SECV). In addition, the residual prediction deviation (RPD), which can be defined as the ratio of the standard deviation (SD) of the reference values over the calibration standard error (RPD = SD/SEC), was calculated and used to evaluate the practical usefulness of predictive models (Hruschaka, 2001 ). Finally, the method was tested by applying the calibration equations obtained to the validation set of samples, which allows the predicted values to be compared with the actual data. The standard error of prediction corrected (SEPC) and the root mean square error (RMSE) statistics were obtained; these are indicative of the prediction. The predicted values were compared with the reference values using Student's t-test for paired values. Calibration and validation of the models was performed using the WinISI Version 4.10 (Infrasoft International, State Collee, PA, USA). 2.4.2 Artificial Neural Network Calibration The neural network models were built with MatLab (R2023a, MathWorks Inc.) as a multilayer feed-forward network. As in the case of the MPLS calibration, the input data were the 451 spectral data recorded in the NIR region and the output data were the quantified percentages of each fatty acid and their corresponding sums of the 148 samples analyzed. The ANN architecture can be described as having an input layer, a hidden layer with a variable number of neurons, and a single neuron in the output layer where the value to be predicted has been obtained. The transfer functions implemented were the sigmoid hyperbolic tangent function for the neurons in the hidden layer and the rectified linear unit (RLU) function for the neuron in the output layer respectively. In total, hundreds of architectures were analyzed with a variable number of neurons in the hidden layer (between 1 and 25) for each parameter, each of which was trained 300 times starting from initial weights randomly assigned from a known seed value, which allows the results to be reproduced (Pillonel et al., 2005 ). The cost function reduction (MSE) was tested with the Scaled Conjugate Gradient, Descendent Gradient, Descendent Gradient with Variable Learning Rate, Descendent Gradient with Momentum, Bayesian Regularization, Quasi-Newton and Levenberg-Marquardt algorithms. The analytical data pairs, i.e. the NIRS data and fatty acid percentages, were randomly divided into three sets to train the network. The first set was a training set (70% of the samples), the second a validation set (15% of the samples), and the third a test set to measure the goodness of the network (15% of the samples). The ANN architectures were then optimized for each of the fatty acids and their summations. The predictive ability of the models was determined by a higher RSQ and a lower RMSE. 3. Results and discussion 3.1 Fatty acid composition The lipid composition of the pieces analyzed both at an intermediate curing point and in the final product is presented in Table 1. The results show a wide variability, which is of interest when pursuing the development of calibration equations. A total of 31 fatty acids were identified and quantified. Of these, the most abundant were monounsaturated fatty acids (MUFA) with percentages between 59.59 and 62.06% with a significantly higher content in the group of intermediate cured hams (20 months). The sum of saturated fatty acids (SFA) presented values of between 31.42 and 32.06 with no significant differences between the four products, while the values of polyunsaturated fatty acids (PUFA) varied between 6.52 and 8.76% with higher values found in the shoulders than in the hams and higher values in the final products than in the intermediate samples. The mean values of the different summations were similar to those obtained by other studies for Iberian ham from animals fattened with the same production system ( montanera ) and the same breeds (100% and 50% Iberian) (Pajuelo et al., 2022). For the sum of the n3, significant differences were found in the four groups analyzed with significantly higher values in the products at the end of the curing period. In the case of the n6, the shoulders showed significantly higher values than the hams in both samples. Table 1. Mean values (± standard deviation) of the fatty acids analysed in shoulder and ham at the intermediate point of curing and in the final product, expressed as % by weight of total fatty acids. Shoulder 8 months N=16 Ham Shoulder Ham 20 months 25 months 36 months N=12 N=60 N=60 Mean Sd Mean Sd Mean Sd Mean Sd P-valor C12:0 0,12 ±0,08b 0,07 ±0,01a 0,09 ±0,08ab 0,09 ±0,02ab 0,099 C14:0 1,41 ±0,36a 1,12 ±0,10a 1,81 ±0,65b 1,34 ±0,15a 0,000 C15:0 0,01 ±0,01 0,02 ±0,00 0,10 ±0,67 0,03 ±0,01 0,764 C16:0 21,67 ±1,70 20,91 ±1,11 22,01 ±1,73 21,91 ±1,26 0,130 C17:0 0,13 ±0,07 0,14 ±0,02 0,13 ±0,30 0,18 ±0,02 0,439 C18:0 8,18 ±1,61ab 8,72 ±0,94b 7,33 ±1,31a 8,23 ±0,97ab 0,000 C20:0 0,15 ±0,09b 0,18 ±0,03b 0,09 ±0,08a 0,17 ±0,04b 0,000 C21:0 0,03 ±0,03a 0,25 ±0,05c 0,04 ±0,04a 0,07 ±0,01b 0,000 C22:0 0,01 ±0,01 0,01 ±0,01 0,02 ±0,05 0,03 ±0,01 0,050 C23:0 0,33 ±0,85b 0,01 ±0,02a 0,13 ±0,33ab 0,00 ±0,01a 0,006 C24:0 0,02 ±0,05b 0,00 ±0,00a 0,01 ±0,02ab 0,02 ±0,01b 0,009 SFA 32,06 ±2,55 31,42 ±1,82 31,67 ±2,24 32,06 ±2,10 0,667 C14:1 n5 0,01 ±0,01 0,02 ±0,01 0,03 ±0,13 0,03 ±0,01 0,806 ∑ C16:1 4,81 ±1,00ab 4,27 ±0,62a 5,24 ±1,37b 4,53 ±0,58ab 0,001 C17:1 0,19 ±0,08 0,21 ±0,03 0,17 ±0,11 0,19 ±0,03 0,275 C18:1 n9t 0,14 ±0,10 0,04 ±0,09 0,36 ±1,15 0,23 ±0,04 0,435 C18:1 50,23 ±2,75a 52,92 ±1,33b 49,19 ±2,69a 50,68 ±1,61a 0,000 C18:1 n7 4,48 ±0,28ab 4,18 ±0,51a 4,35 ±0,45ab 4,66 ±0,48b 0,000 C20:1 n9 0,06 ±0,06 0,15 ±0,02 0,10 ±0,31 0,05 ±0,03 0,265 C22:1 n9 0,36 ±0,10c 0,24 ±0,02b 0,31 ±0,14bc 0,10 ±0,03a 0,000 C24:1 n9 0,00 ±0,01a 0,06 ±0,01b 0,02 ±0,04a 0,01 ±0,03a 0,000 MUFA 60,30 ±2,33a 62,06 ±0,95b 59,59 ±2,09a 60,47 ±1,53a 0,000 C18:2 n6t 0,14 ±0,07b 0,17 ±0,02b 0,11 ±0,10b 0,02 ±0,02a 0,000 C18:2 n6 5,75 ±0,82b 4,81 ±1,23a 6,80 ±0,88c 5,38 ±0,90ab 0,000 C18:3 n6 0,01 ±0,01a 0,36 ±0,12b 0,01 ±0,01a 0,02 ±0,01a 0,000 C18:3 n3 1,27 ±0,26b 1,01 ±0,07a 1,45 ±0,15c 1,33 ±0,12b 0,000 C20:2 n6 0,23 ±0,08b 0,07 ±0,01a 0,19 ±0,13b 0,25 ±0,04b 0,000 C20:3 n6 0,03 ±0,04 0,07 ±0,02 0,05 ±0,11 0,06 ±0,01 0,323 C20:3 n3 0,00 ±0,00a 0,00 ±0,00a 0,01 ±0,08a 0,25 ±0,05b 0,000 C20:4 n6 0,00 ±0,00 0,04 ±0,05 0,12 ±0,62 0,08 ±0,03 0,679 C22:2 n6 0,22 ±0,57 0,00 ±0,01 0,11 ±0,36 0,04 ±0,17 0,143 C20:5 n3 0,00 ±0,00 0,00 ±0,00 0,01 ±0,07 0,00 ±0,00 0,512 C22:6 n3 0,00 ±0,00a 0,00 ±0,01a 0,02 ±0,08a 0,06 ±0,02b 0,000 PUFA 7,64 ±1,23b 6,52 ±1,34a 8,76 ±1,04c 7,45 ±1,10b 0,000 n3 1,28 ±0,26b 1,02 ±0,07a 1,48 ±0,19c 1,64 ±0,15d 0,000 n6 6,68 ±1,26b 5,41 ±1,22a 7,28 ±0,96b 5,81 ±0,97a 0,000 The number of fatty acids quantified differed depending on the type of muscle and stage of maturation; it was found that this number was higher in samples from the Biceps femoris (ham). In all cases it was possible to quantify the characteristic and main fatty acids of Iberian pig fat. Among them are oleic acid (C18:1) with values of between 49.19 and 52.91%, palmitic acid (C16:0) with values of between 20.91 and 22.01%, stearic acid (C18:0) with values of between 7.33 and 8.72%, and linoleic acid (C18:2 n6) with values of between 4.81 and 6.80%. The results obtained coincide with those reported for ham from animals which have fed extensively on acorns and grass (Martín-Cáceres, 1996). In addition to these major fatty acids, it was possible to quantify other fatty acids such as palmitoleic (C16:1), trans-vaccenic (C18:1 n7), myristic (C14:0) and α-linolenic (C18:3 n3) acids which presented lower values between 1.01 and 5.24%. With regard to the effect of the curing time, significant differences were observed in 12 out of the 31 fatty acids quantified between the intermediate sampling (20 months) and the final sampling of the hams (36 months). In the case of the shoulders sampled at 8 and 25 months, these differences were only significant in 4 out of the 31 fatty acids quantified. This fact is due to the processes of lipolysis and oxidation during this processing period, which in turn are influenced by different factors such as the length of the curing period, the level and manner of salting, and the environmental conditions during maturation (Buscailhon & Monin, 1994; Toldrá & Flores, 1998). Furthermore, they may also depend on the type of muscle fiber, as this influences changes in the oxidative stability of the meat, or on lipid oxidation phenomena during storage and processing (A. I. Andrés et al., 2001) which are related to the presence of prooxidant and antioxidant factors (Aristoy & Toldrá, 1998; Henckel, Oksbjerg, Erlandsen, Barton-Gade & Bejerholm, 1997). As for the effect of maturation on the different fatty acids, this was similar in both shoulders and hams. Myristic (C14:0), palmitic (C16:0), palmitoleic (C16:1), linoleic (C18:2 n6), and α-linolenic (C18:3 n3) fatty acids and the sum of SFA and PUFA n3 showed an increase during maturation in both hams and shoulders and these differences were significant for both C14:0 and C18:2 n6 in the case of shoulders. In contrast, a decrease was observed in the percentage of other fatty acids such as C18:0, C20:0, C21:0, C23:0, C18:1, C22:1 n9, C18:2 n6t, and C18:3 n6 and also in the sum of MUFA. In this case the differences were generally significant for hams except for C20:0 for which shoulders showed a significant decrease owing to maturation. In general, it was therefore observed that maturation involved a decrease in MUFA and a significant increase in PUFA. The same behavior observed in this study for C16:0, C16:1, and C18:0 was also reported by Cava, Estévez, Morcuende & Antequera, (2003) in the Semimembranosus muscle. Antequera et al. (1992) also observed a decrease in certain unsaturated triglycerides during the curing process of ham mainly concerning lipolysis and direct oxidation phenomena. 3.2 Spectral characteristics The mean spectra of each of the 148 samples in the wavelength range of 1100 to 2000 nm were similar in type. Although a wide dispersion of the spectra could be observed (Figure 1a), the averages calculated according to product type and maturation time were very close to each other with the exception of the average for 36-month-old ham, which had higher absorbance values. This fact is related to physical factors owing to fibrillar changes caused by proteolysis processes, which are more evident in pieces with a longer curing time. The same behavior has been observed in fresh beef and lamb samples (S. Andrés et al., 2007; Prieto, Andrés, Giráldez, Mantecón & Lavín, 2008; Ripoll, Albertí, Panea, Olleta & Sañudo, 2008). For a better interpretation of the absorption zones, Figure 1b shows the pre-treated spectrum with an initial derivative (1,1,1,1), in which it can be seen that no anomalous spectra or spectral noise appeared throughout the measurement range (Hruschaka, 2001). This representation of the spectrum allows the identification of the absorption maxima and the bands in which the variance is more marked (Zhou, Wu, Li, Wang & Zhang, 2012). These maxima were reached in the second and first overtone bands and are associated with fat owing to the presence of C-H bonds at the 1200, 1730, and 1770 nm wavelengths (Barlocco, Vadell, Ballesteros, Galietta & Cozzolino, 2006; Leroy et al., 2004; Prieto et al., 2008). Other absorption maxima were located in the first and second overtone at 1458 and 1824 nm which correspond to O-H and C-O bonds. The maximum dispersion was located in the bands related to water absorption (1140-1150 nm; 1300-1450 nm), to fats in the second overtone (1165-1200 nm; 1210-1270 nm), to the absorption of fat, water, and protein in the combination bands, (1300-1450 nm) and to the absorption of fats and aromatic compounds in the first overtone (1640-1725 nm; 1734 nm; 1770 nm). The 1680 nm band which indicates the presence of double bonds corresponding to unsaturated fatty acids is of special interest (Garrido-Varo, Carrete & Fernández-Cabanás, 1998). 3.3 Fatty acid calibration equations In order to develop the equations, the fatty acids which could be quantified in at least 90% of the samples were selected. According to this criterion, a total of 15 constituents were selected corresponding to 10 individual fatty acids and the different summations according to the number of unsaturation and their position, i.e. the summations of saturated (SFA), monounsaturated (MUFA), polyunsaturated (PUFA), and the summations of PUFA n3 and PUFA n6 fatty acids. 3.3.1 Calibration equations by the MPLS method As described above, for the development of equations using MPLS a PCA was performed to eliminate possible spectral outliers according to the H > 3.5 criterion. The total set of samples was then divided into a calibration set and an external validation set consisting of 80% (118 samples) and 20% (30 samples) respectively. Potential outliers by the reference method were also removed according to the T > 2.5 criterion. For each constituent a variable number was eliminated and equations which eliminated more than 10% of the samples were not considered valid; this maximum percentage of outlier samples was similar to that used in other studies (Fernández-Cabanás et al., 2011). Based on this criterion, equations could be developed for 13 of the constituents, i.e. 8 fatty acids and the different summations. For each constituent the model was optimized by applying different mathematical and scatter treatments (None, SNV and DT) to obtain the maximum information provided by the chemical signals. It should be noted that the best calibration results were obtained with the mathematical pre-treatments applying a first and second derivative, which is in agreement with other studies related to the prediction of fatty acids in fats and products deriving from Iberian pork (Fernández-Cabanás, Garrido-Varo, García-Olmo, De Pedro & Dardenne, 2007, Fernández-Cabanás et al., 2011). During the development of the equations the model was evaluated by cross-validation. In this case the total number of samples was divided into four sets, using three for calibration and one for prediction; the process was repeated as many times as the number of sets available so that all the sets formed part of the calibration and prediction. This process allowed the validation of the model by checking the predictive capacity by means of the SECV statistic. Of the total equations, 8 presented RSQ values of > 0.5 and RPD values of between 1.41 and 1.93 and corresponded to those fatty acids or summations which are present in the samples at levels higher than 1% (1.33-60.02%) and with a higher variability coefficient (4.37-22.30%) (Table 2). These constituents were the sum of the total PUFA, n3, and n6 and the fatty acids C14:0, C18:0, C18:1, C18:2 n6 and C18:3 n3. For the fatty acids with mean values of < 1% no equations could be developed. The prediction models obtained with RSQ values of between 0.30 and 0.69 can be used to separate samples with higher and lower analytical values, in this case the fatty acids C14:0, C16:0, C18:0, C16:1, C18:1, C18:1 n7, C18:2 n6, SFA, MUFA, and n6. RSQ values above 0.70 would indicate good predictive ability as is the case for C18:3 n3, PUFA, and n3 (Shenk & Westerhaus, 1996). It can also be observed that the determination coefficients of the summations are influenced by the calibration statistics of their main individual fatty acids, which has also been reported in other studies (Fernández-Cabanás, Polvillo, Rodríguez-Acuña, Botella & Horcada, 2011). The best calibration results were thus obtained for polyunsaturated fatty acids and therefore better results are obtained in summates of this group (Table 2). The SECV values obtained in most cases were higher than those determined in other studies predicting fatty acids in meat and meat products made from pork (Fernández-Cabanás et al., 2011; González-Martı́n, González-Pérez, Alvarez-García & González-Cabrera, 2005; Tejerina et al., 2018). Table 2. Descriptors of NIR calibration and results of cross-validation for the Iberian ham and shoulder fatty acid profile by MPLS Fatty acid Mathematic treatment N Terms Mean Range Est. Min Est. Max SD SEC SECV RSQ CV % RPD C14:0 SNV 2,4,4,1 106 6 1.39 0.84-2.42 0.46 2.33 0.31 0.18 0.27 0.67 22.30 1.73 C16:0 DT 2,4,4,1 114 4 21.87 17.20-25.51 17.26 26.47 1.53 1.25 1.44 0.34 6.99 1.23 C18:0 SNV+DT 2,4,4,1 105 6 7.89 5.50-10.20 4.80 10.97 1.03 0.59 0.89 0.67 13.05 1.75 SFA DT 2,4,4,1 113 4 31.98 27.01-37.25 25.46 38.50 2.17 1.75 2.01 0.35 6.78 1.24 C16:1 DT 2,4,4,1 109 3 4.63 2.89-6.61 2.43 6.82 0.73 0.61 0.70 0.30 15.76 1.20 C18:1 DT 1,4,4,1 105 8 50.31 44.85-55.18 43.7 56.93 2.20 1.43 1.77 0.58 4.37 1.54 C18:1 n7 SNV 2,10,10,1 111 8 4.44 3.39-5.48 3.10 5.77 0.45 0.34 0.41 0.43 10.13 1.33 MUFA DT 0,0,1,1 111 9 60.02 54.68-63.93 54.33 65.72 1.90 1.41 1.54 0.45 3.16 1.35 C18:2 n6 None 1,4,4,1 110 8 5.90 3.81-8.18 2.70 9.10 1.07 0.62 0.80 0.66 18.13 1.71 C18:3 n3 SNV+DT 2,4,4,1 105 6 1.33 0.94-1.73 0.84 1.84 0.16 0.09 0.11 0.73 12.03 1.93 PUFA None 2,4,4,1 111 7 8.06 5.72-10.88 4.23 11.89 1.28 0.70 1.05 0.70 15.88 1.84 n3 SNV 2,4,4,1 111 8 1.62 1.2-2.07 1.04 2.21 0.20 0.11 0.17 0.71 12.34 1.85 n6 None 2,10,10,1 112 8 6.33 3.32-9.13 2.91 9.74 1.14 0.81 1.07 0.50 18.00 1.41 Est. Min= Estimated minimum; Est. Max= Estimated maximum; SD= Standard Deviation; SEC= Standard Error of Calibration. SECV: Standard Error of Cross-validation; RSQ= Coefficient of determination; CV= Coefficient of variation; RPD= Ratio of Performance to Deviation (SD/SEC) The MPLS calibration allows the determination of which wavelengths are the most important in the generation of the predictive equations because they are multiplied by a beta coefficient (β) with a higher value (Lucas, Andueza, Ferlay & Martín, 2008; Martínez-Martín, Hrenández-Jiménez, Revilla y Vivar-Quintana, 2023). For each of the fatty acids therefore, the wavelengths with the highest β values were selected and correlated with the bonds and functional groups (information provided by WinISI software) which make the greatest contribution to absorbance at that wavelength (Table 3). In general, the wavelengths with the highest contribution are associated with molecules containing C-H, N-H, O-H, C=O, and S-H bonds with molecular vibrational modes mainly in the first overtone and to a lesser extent in the combination and second overtone bands. Table 3. Spectral wavelengths and chemical groups with high β coefficients in the development of calibration equations Fatty acid Wavelength (nm) Chemical bonds Chemical Structure and Funcional Groups C14:0 1414 O-H, C-H ROH-H2O, CH, CH Oil 1442 C-H CH2 1452 O-H Water 1460 N-H Amides 1480 N-H Amides 1922 C=O, O-H -CONH C16:0 1470 N-H Amides 1480 N-H Amines 1922 C=O, O-H -CONH C18:0 1454 O-H Water 1472 N-H Amides 1496 N-H, O-H Amides 1540 O-H Water SFA 1506 N-H Protein 1518 N-H Urea 1542 O-H Water C16:1 1480 N-H Glucose 1520 N-H Urea 1574 N-H -CONH C18:1 1410 O-H, C-H Water, C-H Oil 1722 C-H C-H Oil 1746 S-H S-H 1800 C-H -CH2 1918 C=O -CONH C18:1 n7 1364 C-H -CH3 1534 N-H -RNH2 1622 C-H =CH2 MUFA 1494 N-H, O-H, N-H -CONHR, ARNH2, Urea 1902 C=O -CO2H 1920 C=O -CONH C18:2 n6 1166 C-H HC=CH 1442 C-H, O-H -CH2, Aromatic compounds 1510 N-H Protein 1614 O-H, C-H ROH-Water, -CH, Aromatic compounds 1938 O-H Water C18:3 n3 1454 O-H -STARCH 1568 N-H -CONH 1620 C-H =CH2 1816 O-H -Cellulose PUFA 1444 C-H, O-H -CH2, Aromatic compounds 1460 O-H, N-H -CONH2 1496 N-H, O-H -CONHR, -ARNH2 1568 N-H -CONH 1622 C-H =CH2 n3 1442 C-H CH2 1494 N-H, O-H -CONHR, -ARNH2 1566 N-H -CONH 1796 O-H -WATER n6 1576 N-H -ONH 1770 C-H -CH2 1954 C=O -CO2R It can be observed that saturated fatty acids were strongly correlated with the 1452 and 1540 nm wavelengths which correspond to O-H bonds in the first overtone region, and also with the 1460-1520 nm wavelengths which correspond to N-H bonds in the first overtone region. Monounsaturated fatty acids showed the most significant β values at 1410 nm, a wavelength related to O-H bonds, and at wavelengths of 1480 to 1580 nm related to N-H bonds in the first overtone region. Other areas where coefficients with high values were observed were the wavelengths between 1166-1365 nm and 1600-1800 nm related to C-H bonds. High value coefficients appeared at wavelengths of above 1900 nm correlated with the C=O bond in the second overtone region. It should be noted that in the case of the MUFA summation calibration the last three bonds (C-H, C-O, and C=O) have the highest weight. In the case of PUFA, the β coefficients with the highest values were at wavelengths 1166, 1442, 1620, and 1770 associated with the C-H bonds. Other coefficients appeared at wavelengths 1440-1460, 1614, 1800, and 1938 nm in association with O-H bonds and at wavelengths of between 1500 and 1580 nm associated with N-H bonds in the first overtone region. In addition, it should be added that at 1954 nm there was a high β coefficient value for the omega 6 calibration related to the C=O bond located in the second overtone region. Although several studies have addressed the prediction of the lipid profile and the sums of the different groups of total fatty acids according to the amount of unsaturation through NIR spectroscopy in fresh beef, pork and rabbit meat (González-Martı́n et al., 2005; Pla, Hernández, Ariño, Ramírez & Díaz, 2007; Realini, Duckett & Windham, 2004) and in subcutaneous fat from Iberian pork (González-Martín et al., 2021; Pérez-Marín, De Pedro Sanz, Guerrero-Ginel & Garrido-Varo, 2009) and obtained acceptable equations, previous research on the prediction of the lipid profile in cured products deriving from Iberian pork is scarce (Fernández-Cabanás et al., 2011; Tejerina et al., 2018). The results of Tejerina et al. (2018) revealed that the salt content and palmitic and oleic fatty acid percentages could be predicted with RSQ values of > 0.8 in packaged Iberian ham. Furthermore, Fernández-Cabanas et al. (2011) evaluated the feasibility of NIRS technology to determine the fatty acid profile in cured sausages from the Iberian pig ( salchichón and chorizo ); they obtained RSQs of between 0.77 and 0.94 for the major fatty acids (C16:0, C18:0, C18:1, C18:2, and C18:3) and for the sums by degree of unsaturation. As for the minority fatty acids, although it was possible to calibrate some of them such as C12:0, C14:0, C17:0, C17:0, C17:1, and C20:1 with RSQ values of between 0.62 and 0.82 in fresh intact Iberian pork loin (González-Martı́n et al., 2005), other studies have however reported determination coefficients of nearly zero for the calibrations developed in dry-cured sausages for minor constituents such as C12:0, C17:0, C17:0, C17:1; C20:0, and C20:1 (Fernández-Cabanás et al., 2011) or coefficients lower than 0.5 in minced and homogenized rabbit meat (Pla et al., 2007). In general, the recording of intact samples presents limitations as the models are less accurate in homogenized and minced samples (Cozzolino & Murray, 2002) owing to the different organization of muscle fibers, physical and chemical characteristics, the meat cut, the intramuscular fat content, and the moisture among other factors (Cozzolino, Murray, Scaife & Paterson 2000; Cozzolino & Murray, 2002). The correlation of the values obtained in the laboratory with regard to those predicted by NIR with the fiber optic probe for the calibration set has been plotted in Figure 2 for stearic, α-linolenic, oleic, linoleic, AGP, and n3 fatty acids. It can be seen that the validation RSQ values were higher than 0.6 and that the SEP and SEP (C) values were similar. The graphs allow the observation of the linear relationship between the reference values and those predicted in the calibration set. As far as we know, there are no previous studies on the prediction of fatty acids in hams or shoulders with which to compare RSQ values. In order to check the robustness of the model, the equations obtained were applied to 30 samples other than those used for calibration for which the spectral record and analytical data of the lipid profile are available. This allowed comparison of the predicted values with reference data obtained by laboratory analysis. The prediction results for the validation set of samples confirmed that there were no significant differences between the predicted data and the laboratory value (p>0.05) for all fatty acids except for myristic acid (C14:0). The root mean standard error (RMSE) values were between 0.13 for (C18:3 n6) and 2.02 for the summation of MUFA. 3.3.2 Calibration equations by the ANN method As previously described, the first step for developing calibration equations using artificial neural networks was to test what was the most suitable algorithm by using the cost function reduction criterion. The algorithm optimization process determined that the Scaled Conjugate Gradient, the Gradient Descent, the Gradient Descent with Adaptive Learning Rate, and the Gradient Descent with Momentum did not provide good predictive results. Better results were obtained with the Bayesian Regularization and Quasi-Newton algorithms albeit at a very high computational time cost. Finally, the Levenberg-Marquardt (LM) algorithm was used as it allowed a greater number of networks with RSQ values above 0.7 and at a reasonable time cost. This same algorithm obtained the best results in the prediction of sensory parameters for raw cured products (Hernández-Jiménez et al., 2020; Hernández-Ramos et al., 2020). The ANN was then optimized to determine the number of neurons in the hidden layer, which had to be between the size of the input and output layer and was determined empirically (Berry & Linoff, 1997; Boger & Guterman, 1997). For each parameter, architectures with a variable number of neurons in the hidden layer (between 1 and 25) were analyzed and each of the architectures was trained 300 times from initial weights randomly assigned from a known seed value. The optimal number of neurons in the hidden layer depended on the parameter or constituent, which means that a higher number of neurons was necessary for the prediction of saturated fatty acids while the rest of the constituents presented a variable number of neurons, highlighting the value of less than 5 neurons in oleic, erucic, linoleic, and PUFA fatty acids (Table 4). Table 4. Number of neurons in the hidden layer, RSQ, Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) values of the best ANN architecture for ecah fatty acid and summatories Number of neurons in the hidden layer Networks ≥0.8 Networks ≥0.7 RSQ MSE RMSE C14:0 20 0 11 0.63 0.10 0.31 C16:0 30 0 11 0.63 1.01 1.00 C18:0 14 0 10 0.57 0.77 0.86 SFA 25 0 23 0.62 1.93 1.39 ∑ C16:1 2 0 5 0.57 0.49 0.70 C18:1 4 7 234 0.70 2.08 1.44 C18:1 n7 29 0 3 0.56 0.12 0.35 C22:1 n9 1 267 1721 0.81 0.00 0.07 MUFA 13 0 111 0.63 1.54 1.24 C18:2 n6 4 17 267 0.73 0.39 0.62 C18:3 n3 24 84 1080 0.76 0.01 0.10 C20:2 n6 23 0 2 0.56 0.00 0.07 PUFA 2 3 177 0.68 0.56 0.75 n3 22 208 1507 0.76 0.02 0.12 n6 22 8 162 0.70 0.56 0.75 RSQ: Coefficient of Determination, MSE: Mean Square Error, RMSE: Root Mean Square Error Furthermore, to perform this network optimization the total set of samples was divided into three sets; i.e. a training set with 104 samples, the validation set with 22 samples, and the test set with 22 samples. For each parameter individual training was carried out to obtain the predictive neural network as detailed above. During the training of the network, it has to be checked that overtraining (or overfitting) of the network is not taking place; in order to do so the 22 samples of the validation set which had not participated in the training were used. Finally, the performance of the network was evaluated with the remaining 22 samples not included in the training and validation sets so as to obtain the RSQ and RMSE values and thus allow the selection of the best network. The results show that in all cases RSQ values of > 0.5 were obtained with the figures ranging between 0.56 and 0.81. The correlation coefficients in the case of lipid profile prediction in rapeseed oil were slightly higher with RSQ values of > 0.7 (Rajković et al., 2023). For all summations (SFA, MUFA, PUFA, n6, and n3) the RSQ values were higher than 0.6 (Table 4). It should be noted that the fatty acids and summations which presented the highest RSQ coefficients, with values of between 0.7 and 0.81, were the fatty acids C18:1, C18:2 n6, C18:3 n3, C22:1 n9, and n3, which showed a higher correlation with the free-range feeding of Iberian pigs (Hernández-Jiménez et al., 2021). A parameter which can be taken into account to ascertain the capacity of the ANN to predict the components analyzed is the number of networks which give an RSQ value higher than that established as a reference (Hernández-Ramos et al., 2020). In this case it was only possible to find networks with an RSQ value of > 0.65 for oleic, erucic, β and α-linolenic, PUFA, n3 and n6 acids, and for some of them the number of networks is less than 10 which gives an idea of the difficulty of calibrating these components. Although it has previously been reported that the higher the number of neurons in the hidden layer and the number of training sessions the better the predictive capacity of the network (Hernández-Ramos et al., 2020), the results obtained in this study showed that there was no direct relationship between the number of neurons in the hidden layer and the RSQ. The RMSEs showed values of between 0.065 and 1.444 with errors of less than 1 in the prediction of all polyunsaturated fatty acids and their respective summations. Lower RMSE values were observed for the prediction of C18:1, C22:1, and C18:3 n3 and higher values for C16:0, C18:0, and C18:2 fatty acids than those reported in other studies predicting the lipid profile by ANN (Rajković et al., 2023). Finally, Figure 3 shows the prediction plots and RSQ values for some of the calculated fatty acids and sum values. They show a satisfactory correlation between the reference values and the values predicted by the network. 3.4 Comparison of the MPLS and ANN results If the results obtained by the MPLS and ANN methods are compared it can be seen that by using neural networks, it was possible to predict the 15 constituents by applying the RSQ > 0.5 criterion; by MPLS it was only possible to predict 8 constituents under this criterion. Furthermore, the use of ANNs allowed the obtaining of 4 equations with an RSQ value of > 0.7 while with the application of the MPLS methodology it was possible to predict only 3 parameters with this criterion (Tables 2 and 4). The performance of ANN was significantly better compared with the multiple regression model (Fig 2 and Fig 3), which is probably due to the fact that it is a non-linear method. This result is in line with previous studies conducted for the prediction of sensory parameters of meat products (Hernández-Jiménez et al., 2020), fruit firmness on harvesting (Torkashvan, Ahmadi & Nikravesh, 2017), and cheese texture during the ripening process (Vásquez et al., 2018). However, other authors have reported similar results in the predictive ability of the physicochemical parameters of fresh pork using ANN and MPLS (Prevolnik, Čandek-Potokar, Novič.& Škorjanc, 2009). The wavelengths which have had the greatest weight in the generation of the models by the MPLS and ANN regression methods were compared. The interpretation is simpler in the case of the MPLS regression method since according to the general linear model NIRS data (Y) can be modelled as Y = b 0 + b 1 C l1 + b 2 C l2 + b 3 C l3 +……+ b n C ln , in which β refers to the coefficients measuring the contribution of each wavelength to explaining Y and C l1 ,C l2 ,C l3 ,....C ln , are the wavelengths. A higher β value therefore means a higher contribution of that wavelength to the final equation. In the case of ANN, the methodology previously described by Hernández-Ramos et al. (2020) was applied. The effect of each input wavelength on the modelled variables was assessed by analyzing the weight of each of the inputs for each of the neurons in the hidden layer, as this value could reflect the importance of that input in the correct classification or predictive ability of that neuron. Once all the weights (positive or negative) were calculated for each of the neurons in the hidden layer (the number of which depended on the predicted variable), the quadratic weight of each of the inputs was calculated, i.e. the sum of the squared weight of each of the n neurons in the hidden layer for each of the 451 wavelengths. This process was carried out for all fatty acids and the data obtained for some of the most important ones (C16:0, C18:1, and C18:2 n3) are shown in Figure 4. To compare the MPLS and ANN results in a simple way, the square of each of the β values obtained for each of the 451 wavelengths was also calculated and plotted for these three fatty acids. First of all, for the MPLS calibration it can clearly be seen that it is not a single wavelength as shown in Table 3 but a band around those β which make a significant contribution to the calibration. In the case of the C16:0 fatty acid, the following bands can be observed for the MPLS method with major contributions located around 1170-1200, 1260-1280, 1460-1480, 1510, 1580, 1610, 1660, 1780-1800, and 1910-1930. As mentioned above, these would correspond to absorption bands in the first and second overtones associated with the C-H, N-H, O-H and C=O chemical bonds. For this fatty acid it was observed that there was a coincidence with the inputs with more weight for ANN in the wavelengths close to 1180, 1260, and 1650, which correspond to the absorption bands of C-H bonds, and the 1510 band which corresponds to the absorption of N-H bonds. In addition to these wavelengths, when ANN calibration was carried out high quadratic weights were observed for the 1100 to 1300 bands, which are related to C-H and N-H bonds, for the 1380 to 1440 range in which the main absorbing bonds are C-H and O-H, for the wavelengths between 1540 and 1590 which are associated with O-H bonds, and for the 1660-1760 range corresponding to the first overtone of the C-H bond. The C-H, N-H and O-H bonds were therefore the most relevant for the calibration of this fatty acid as was previously observed in the case of MPLS. As far as C18:1 fatty acid is concerned, in addition to the bands around the β values previously indicated for the calibration with MPLS at 1410, 1722, 1746, and 1918 nm (Table 3), which are also important in calibration with ANN, this graph allows us to observe that there were other relevant bands in the range of 1110 to 1120 nm and around 1200 nm which correspond to the absorbance of C-H and O-H bonds and in the 1260-1280 interval. In addition to these, the bands around 1380 to 1590, 1870 to 1900, and 1940 to 1960 are of great importance in ANN calibration. These bands correspond to the absorbance of C-H, O-H, and N-H bonds in the first interval, of O-H and C=O bonds in the second interval, and of O-H, C=O, and N-H bonds in the last of the above-mentioned regions. These results confirm what was previously observed in MPLS, i.e. the important contribution of C=O bonds in the calibration of unsaturated fatty acids. Finally, Figure 4c presents the results for the C18:3 n6 to show a smaller number of β-values which are relevant in MPLS calibration and narrower bands in both methods. In the MPLS method it can be observed that the bands with the highest contribution to the model development are located at 1170, 1440, 1500 to 1510, and 1780 as already discussed, and also that there is a significant contribution from the bands at 1120, 1210, and again in the 1260-1280 range, all of which are associated with the first and second overtone of the C-H bond. These bands do not correspond to those of greater weight in the ANN method, in which the most important contribution was distributed throughout the entire spectral range between 1380 and 2000 nm with narrow bands. The contributions found at 1610, 1720, 1820, and 1910 nm stand out, which as indicated in Table 3 correspond to the first overtone of C-H, to the C-O bond in oils, to the O-H bond, and to the C=O bond respectively. It should therefore be noted that in both methods the wavelengths with the highest weights are associated with the C-H, N-H, and O-H bonds, together with the C=O bond for unsaturated fatty acids. However, the coefficients in the MPLS method follow a distribution in defined bands over the whole spectral range, while in the case of ANN their behavior was very different depending on the component to be calibrated. Thus in some of the cases the highest weights appeared in the form of defined bands such as for oleic and linoleic acids. It should be noted that these fatty acids had a small number of neurons in the hidden layer (4), which probably explains this discrete distribution. However, for others such as palmitic acid weights are generally higher and more homogeneously distributed throughout the spectral range, which could be associated with the higher number of neurons in the hidden layer. Finally, the different assignment of weights to the different wavelengths, based on the different algorithms used, is probably the reason for the higher predictive ability of ANNs. 4. Conclusion The results obtained in this study indicate that it is possible to predict the main fatty acids and their summations calculated according to the degree of unsaturation in samples of Iberian hams and shoulders using non-destructive and rapid techniques such as NIR spectroscopy, which would facilitate the nutritional evaluation of the products and their labelling. Of the two chemometric methods used, artificial neural networks were able to predict a greater number of parameters with higher RSQ values than the MPLS method. These results indicate that in the prediction of the percentage of fatty acids in a complex matrix such as cured Iberian pork non-linear methods such as neural networks provided better results. However, this research could be continued by extending it with a larger number of samples in order to generate more robust models and confirm the behavior observed for both predictive methods. Declarations Competing interest Financial interests: The authors declare they have no financial interests. Funding: This research was funded by Salamanca County Council (Spain) under grant number 18VEUH 463AC06. Author Contribution Miriam Hernández-Jiménez and Pedro Hernández Ramos: formal analysis, investigation, visualization. Isabel Revilla conceptualization, supervision, project administration, funding acquisition. Ana Vivar-Quintana: funding acquisition, conceptualization, editing. All the authors wrote and reviewed the manuscript. Acknowledgement We are grateful to “Carrasco Ibéricos” Guijuelo (Salamanca) for their collaboration and Hernández-Jiménez M. is grateful to the Own Program III: Grants for Pre-doctoral Contracts of the University of Salamanca co-funded by Banco Santander. References Alhamdan, A. M., Fickak, A., & Atia, A. R. (2019). Evaluation of sensory and texture profile analysis properties of stored Khalal Barhi dates nondestructively using Vis/NIR spectroscopy. Journal of Food Process Engineering , 42 (6), e13215. https://doi.org/10.1111/jfpe.13215 . Alomar, D., Gallo, C., Castañeda, M., & Fuchslocher, R. (2003). Chemical and discriminant analysis of bovine meat by near infrared reflectance spectroscopy (NIRS). Meat Science , 63 (4), 441–450. https://doi.org/10.1016/S0309-1740(02)00101-8 . Andrés, A. I., Cava, R., Mayoral, A. I., Tejeda, J. F., Morcuende, D., & Ruiz, J. (2001). Oxidative stability and fatty acid composition of pig muscles as affected by rearing system, crossbreeding and metabolic type of muscle fibre. Meat Science , 59 (1), 39–47. https://doi.org/10.1016/S0309-1740(01)00050-X . Andrés, S., Murray, I., Navajas, E. A., Fisher, A. V., Lambe, N. R., & Bünger, L. (2007). Prediction of sensory characteristics of lamb meat samples by near infrared reflectance spectroscopy. Meat Science , 76 (3), 509–516. https://doi.org/10.1016/j.meatsci.2007.01.011 . Ansorena, D., & Astiasarán, I. (2004). Effect of storage and packaging on fatty acid composition and oxidation in dry fermented sausages made with added olive oil and antioxidants. Meat Science , 67 (2), 237–244. https://doi.org/10.1016/j.meatsci.2003.10.011 . Antequera, T., López-Bote, C. J., Córdoba, J. J., García, C., Asensio, M. A., Ventanas, J., García-Regueiro, J. A., & Díaz, I. (1992). Lipid oxidative changes in the processing of Iberian pig hams. Food Chemistry , 45 (2), 105–110. https://doi.org/10.1016/0308-8146(92)90018-W . Arce, L., Domínguez-Vidal, A., Rodríguez-Estévez, V., López-Vidal, S., Ayora-Cañada, M. J., & Valcárcel, M. (2009). Feasibility study on the use of infrared spectroscopy for the direct authentication of Iberian pig fattening diet. Analytica Chimica Acta , 636 (2), 183–189. https://doi.org/10.1016/j.aca.2009.01.058 . Aristoy, M. C., & Toldrá, F. (1998). Concentration of free amino acids and dipeptides in porcine skeletal muscles with different oxidative patterns. Meat Science , 50 (3), 327–332. https://doi.org/10.1016/S0309-1740(98)00037-0 . Balabin, R. M., & Safieva, R. Z. (2011). Near-Infrared (NIR) Spectroscopy for Biodiesel Analysis: Fractional Composition, Iodine Value, and Cold Filter Plugging Point from One Vibrational Spectrum. Energy & Fuels , 25 (5), 2373–2382. https://doi.org/10.1021/ef200356h . Balabin, R. M., Safieva, R. Z., & Lomakina, E. I. (2007). Comparison of linear and nonlinear calibration models based on near infrared (NIR) spectroscopy data for gasoline properties prediction. Chemometrics and Intelligent Laboratory Systems , 88 (2), 183–188. https://doi.org/10.1016/j.chemolab.2007.04.006 . Balabin, R. M., Safieva, R. Z., & Lomakina, E. I. (2008). Wavelet neural network (WNN) approach for calibration model building based on gasoline near infrared (NIR) spectra. Chemometrics and Intelligent Laboratory Systems , 93 (1), 58–62. https://doi.org/10.1016/j.chemolab.2008.04.003 . Barlocco, N., Vadell, A., Ballesteros, F., Galietta, G., & Cozzolino, D. (2006). Predicting intramuscular fat, moisture and Warner-Bratzler shear force in pork muscle using near infrared reflectance spectroscopy. Animal Science , 82 (1), 111–116. https://doi.org/10.1079/ASC20055 . Berry, M. J. A., & Linoff, G. S. (1997). Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management . Wiley. Boger, Z., & Guterman, H. (1997). Knowledge extraction from artificial neural network models. Computational Cybernetics and Simulation 1997 IEEE International Conference on Systems, Man, and Cybernetics , 4 , 3030–3035 vol.4. https://doi.org/10.1109/ICSMC.1997.633051 . Buscailhon, S., & Monin, G. (1994). Déterminisme des qualités sensorielles du jambon sec. Déterminisme Des Qualités Sensorielles Du Jambon Sec , 15 (1), 23–34. Carrapiso, A. I., Bonilla, F., & Garcı́a, C. (2003). Effect of crossbreeding and rearing system on sensory characteristics of Iberian ham. Meat Science , 65 (1), 623–629. https://doi.org/10.1016/S0309-1740(02)00256-5 . Cava, R., Estévez, M., Morcuende, D., & Antequera, T. (2003). Evolution of fatty acids from intramuscular lipid fractions during ripening of Iberian hams as affected by α-tocopheryl acetate supplementation in diet. Food Chemistry , 81 (2), 199–207. https://doi.org/10.1016/S0308-8146(02)00413-2 . Cava, R., Ruiz, J., López-Bote, C., Martín, L., García, C., Ventanas, J., & Antequera, T. (1997). Influence of finishing diet on fatty acid profiles of intramuscular lipids, triglycerides and phospholipids in muscles of the Iberian pig. Meat Science , 45 (2), 263–270. https://doi.org/10.1016/S0309-1740(96)00102-7 . Cozzolino, D., & Murray, I. (2002). Effect of Sample Presentation and Animal Muscle Species on the Analysis of Meat by near Infrared Reflectance Spectroscopy. Journal of Near Infrared Spectroscopy , 10 (1), 37–44. Cozzolino, D., Murray, I., Scaife, J. R., & Paterson, R. (2000). Study of dissected lamb muscles by visible and near infrared reflectance spectroscopy for composition assessment. Animal Science , 70 (3), 417–423. De Maesschalck, R., Jouan-Rimbaud, D., & Massart, D. L. (2000). The Mahalanobis distance. Chemometrics and Intelligent Laboratory Systems , 50 (1), 1–18. https://doi.org/10.1016/S0169-7439(99)00047-7 . Fernández-Cabanás, V. M., Garrido-Varo, A., García Olmo, J., De Pedro, E., & Dardenne, P. (2007). Optimisation of the spectral pre-treatments used for Iberian pig fat NIR calibrations. Chemometrics and Intelligent Laboratory Systems , 87 (1), 104–112. https://doi.org/10.1016/j.chemolab.2006.10.005 . Fernández-Cabanás, V. M., Polvillo, O., Rodríguez-Acuña, R., Botella, B., & Horcada, A. (2011). Rapid determination of the fatty acid profile in pork dry-cured sausages by NIR spectroscopy. Food Chemistry , 124 (1), 373–378. https://doi.org/10.1016/j.foodchem.2010.06.031 . Folch, J., Lees, M., & Sloane Stanley, G. H. (1957). A simple method for the isolation and purification of total lipids from animal tissues. Journal Of Biological Chemistry , 226 (1), 497–509. Gandemer, G. (2002). Lipids in muscles and adipose tissues, changes during processing and sensory properties of meat products. Meat Science , 62 (3), 309–321. https://doi.org/10.1016/S0309-1740(02)00128-6 . García, A. I. M., Andrada, A. D., Muñoz, A. I. R., & Bote, C. J. L. (2005). Effect of feeding level during the period previous to free range fattening on growth and carcass characteristics. Spanish Journal of Agricultural Research , 4 , 387–395. García-Olmo, J., Garrido-Varo, A., & De Pedro, E. (2001). The Transfer of Fatty Acid Calibration Equations Using Four Sets of Unsealed Liquid Standardisation Samples. Journal of Near Infrared Spectroscopy , 9 (1), 49–62. https://doi.org/10.1255/jnirs.294 . Garrido-Varo, A., Carrete, R., & Fernández-Cabanás, V. (1998). Use of Difference near Infrared Reflectance Spectra to Extract Relevant Information from the Spectra of Agro–Food Products. Journal of Near Infrared Spectroscopy , 6 (1), 89–95. https://doi.org/10.1255/jnirs.125 . González, E., & Tejeda, J. F. (2007). Effects of dietary incorporation of different antioxidant extracts and free-range rearing on fatty acid composition and lipid oxidation of Iberian pig meat. Animal , 1 (7), 1060–1067. https://doi.org/10.1017/S1751731107000195 . González-Martín, M. I., Escuredo, O., Hernández-Jiménez, M., Revilla, I., Vivar-Quintana, A. M., Martínez-Martín, I., & Hernández-Ramos, P. (2021). Prediction of stable isotopes and fatty acids in subcutaneous fat of Iberian pigs by means of NIR: A comparison between benchtop and portable systems. Talanta , 224 , 121817. https://doi.org/10.1016/j.talanta.2020.121817 . González-Martín, M. I., Palacios, V., Revilla, V., Vivar-Quintana, I., A. M., & Hernández-Hierro, J. M. (2017). Discrimination between cheeses made from cow’s, ewe’s and goat’s milk from unsaturated fatty acids and use of the canonical biplot method. Journal of Food Composition and Analysis , 56 , 34–40. https://doi.org/10.1016/j.jfca.2016.12.005 . González-Martı́n, I., González-Pérez, C., Alvarez-Garcı́a, N., & González-Cabrera, J. M. (2005). On-line determination of fatty acid composition in intramuscular fat of Iberian pork loin by NIRs with a remote reflectance fibre optic probe. Meat Science , 69 (2), 243–248. https://doi.org/10.1016/j.meatsci.2004.07.003 . Henckel, P., Oksbjerg, N., Erlandsen, E., Barton-Gade, P., & Bejerholm, C. (1997). Histo- and biochemical characteristics of the Longissimus dorsi muscle in pigs and their relationships to performance and meat quality. Meat Science , 47 (3), 311–321. https://doi.org/10.1016/S0309-1740(97)00063-6 . Hernández-Jiménez, M., González-Martín, M. I., Martínez-Martín, I., Revilla, I., & Vivar-Quintana, A. M. (2021). Carbon stable isotopes, fatty acids and the use of NIRS to differentiate IBERIAN pigs. Meat Science , 182 , 108619. https://doi.org/10.1016/j.meatsci.2021.108619 . Hernández-Jiménez, M., Hernández-Ramos, P., Martínez-Martín, I., Vivar-Quintana, A. M., González-Martín, I., & Revilla, I. (2020). Comparison of artificial neural networks and multiple regression tools applied to near infrared spectroscopy for predicting sensory properties of products from quality labels. Microchemical Journal , 159 , 105459. https://doi.org/10.1016/j.microc.2020.105459 . Hernández-Ramos, P., Vivar-Quintana, A. M., Revilla, I., González-Martín, M. I., Hernández-Jiménez, M., & Martínez-Martín, I. (2020). Prediction of Sensory Parameters of Cured Ham: A Study of the Viability of the Use of NIR Spectroscopy and Artificial Neural Networks. Sensors (Basel, Switzerland) , 20 (19). https://doi.org/10.3390/s20195624 . Article 19. Horcada, A., Valera, M., Juárez, M., & Fernández-Cabanás, V. M. (2020). Authentication of Iberian pork official quality categories using a portable near infrared spectroscopy (NIRS) instrument. Food Chemistry , 318 , 126–133. https://doi.org/10.1016/j.foodchem.2020.126471 . Hruschaka, W. (2001). Data analysis: Wavelengt selection methods (P. C. Williams & N. Karl, Eds.). Near infrared technology in the agricultural and food industries, American Association of Cereal Chemist. Jurinjak Tušek, A., Jurina, T., Čulo, I., Valinger, D., Gajdoš Kljusurić, J., & Benković, M. (2022). Application of NIRs coupled with PLS and ANN modelling to predict average droplet size in oil-in-water emulsions prepared with different microfluidic devices. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy , 270 , 120860. https://doi.org/10.1016/j.saa.2022.120860 . Kundu, P., Paul, V., Kumar, V., & Mishra, I. M. (2015). Formulation development, modeling and optimization of emulsification process using evolving RSM coupled hybrid ANN-GA framework. Chemical Engineering Research and Design , 104 , 773–790. https://doi.org/10.1016/j.cherd.2015.10.025 . Leroy, B., Lambotte, S., Dotreppe, O., Lecocq, H., Istasse, L., & Clinquart, A. (2004). Prediction of technological and organoleptic properties of beef Longissimus thoracis from near-infrared reflectance and transmission spectra. Meat Science , 66 (1), 45–54. https://doi.org/10.1016/S0309-1740(03)00002-0 . Lucas, A., Andueza, D., Ferlay, A., & Martin, B. (2008). Prediction of fatty acid composition of fresh and freeze-dried cheeses by visible–near-infrared reflectance spectroscopy. International Dairy Journal , 18 (6), 595–604. https://doi.org/10.1016/j.idairyj.2007.12.001 . Marini, F. (2013). In M. Bevilacqua, R. Bucci, A. D. Magrì, A. L. Magrì, R. Nescatelli, & F. Marini (Eds.), Chemometrics in food chemistry Chap. 5. Classification and class-modelling (Vol. 1,281st ed., 28).). Elsevier Inc. Martínez-Martín, I., Hernández-Jiménez, M., Revilla, I., & Vivar-Quintana, A. M. (2023). Prediction of Mineral Composition in Wheat Flours Fortified with Lentil Flour Using NIR Technology. Sensors (Basel, Switzerland) , 23 (3). Article 3. https://doi.org/10.3390/s23031491 . Martín-Cáceres, L. (1996). Influencia de las condiciones del procesado sobre los cambios madurativos en el jamón ibérico [Tesis docotral, Universidad de Extremadura]. Martín-Gómez, A., Arroyo-Manzanares, N., Rodríguez-Estévez, V., & Arce, L. (2019). Use of a non-destructive sampling method for characterization of Iberian cured ham breed and feeding regime using GC-IMS. Meat Science , 152 , 146–154. https://doi.org/10.1016/j.meatsci.2019.02.018 . Scopus. Mataix, J., Quiles, J. L., & Rodríguez, J. (2001). Aporte de grasa. Guías Alimentarias Para La Población Española Recomendaciones Para Una Dieta Saludable Madrid: SENC , 231–237. Norris, K. H. (2001). Understanding and Correcting the Factors Which Affect Diffuse Transmittance Spectra. NIR News , 12 (3), 6–9. https://doi.org/10.1255/nirn.613 . Ortiz, A., León, L., Contador, R., & Tejerina, D. (2021). Near-Infrared Spectroscopy (NIRS) as a Tool for Classification of Pre-Sliced Iberian Salchichón, Modified Atmosphere Packaged (MAP) According to the Official Commercial Categories of Raw Meat. Foods , 10 (8). https://doi.org/10.3390/foods10081865 . Article 8. Pajuelo, A., Sánchez, S., Pérez-Palacios, T., Caballero, D., Díaz, J., Antequera, T., & Marcos, C. F. (2022). 1H NMR to analyse the lipid profile in the glyceride fraction of different categories of Iberian dry-cured hams. Food Chemistry , 383 , 132371. https://doi.org/10.1016/j.foodchem.2022.132371 . Pérez-Marín, D., De Pedro Sanz, E., Guerrero-Ginel, J. E., & Garrido-Varo, A. (2009). A feasibility study on the use of near-infrared spectroscopy for prediction of the fatty acid profile in live Iberian pigs and carcasses. Meat Science , 83 (4), 627–633. https://doi.org/10.1016/j.meatsci.2009.07.012 . Pérez-Marín, D., Fearn, T., Riccioli, C., De Pedro, E., & Garrido, A. (2021). Probabilistic classification models for the in situ authentication of iberian pig carcasses using near infrared spectroscopy. Talanta , 222 , 121511. https://doi.org/10.1016/j.talanta.2020.121511 . Pérez-Marín, D., Garrido-Varo, A., De Pedro, E., & Guerrero-Ginel, J. E. (2007). Chemometric utilities to achieve robustness in liquid NIRS calibrations: Application to pig fat analysis. Chemometrics and Intelligent Laboratory Systems , 87 (2), 241–246. https://doi.org/10.1016/j.chemolab.2007.02.004 . Pillonel, L., Badertscher, R., Casey, M., Meyer, J., Rossmann, A., Schlichtherle-Cerny, H., Tabacchi, R., & Bosset, J. O. (2005). Geographic origin of European Emmental cheese: Characterisation and descriptive statistics. International Dairy Journal , 15 (6), 547–556. https://doi.org/10.1016/j.idairyj.2004.07.028 . Pla, M., Hernández, P., Ariño, B., Ramírez, J. A., & Díaz, I. (2007). Prediction of fatty acid content in rabbit meat and discrimination between conventional and organic production systems by NIRS methodology. Food Chemistry , 100 (1), 165–170. https://doi.org/10.1016/j.foodchem.2005.09.029 . Prevolnik, M., Čandek-Potokar, M., Novič, M., & Škorjanc, D. (2009). An attempt to predict pork drip loss from pH and colour measurements or near infrared spectra using artificial neural networks. Meat Science , 83 (3), 405–411. https://doi.org/10.1016/j.meatsci.2009.06.015 . Prieto, N., Andrés, S., Giráldez, F. J., Mantecón, A. R., & Lavín, P. (2008). Ability of near infrared reflectance spectroscopy (NIRS) to estimate physical parameters of adult steers (oxen) and young cattle meat samples. Meat Science , 79 (4), 692–699. https://doi.org/10.1016/j.meatsci.2007.10.035 . Prieto, N., Pawluczyk, O., Dugan, M. E. R., & Aalhus, J. L. (2017). A Review of the Principles and Applications of Near-Infrared Spectroscopy to Characterize Meat, Fat, and Meat Products. Applied Spectroscopy , 71 (7), 1403–1426. https://doi.org/10.1177/0003702817709299 . Rajković, D., Marjanović Jeromela, A., Pezo, L., Lončar, B., Grahovac, N., & Kondić Špika, A. (2023). Artificial neural network and random forest regression models for modelling fatty acid and tocopherol content in oil of winter rapeseed. Journal of Food Composition and Analysis , 115 , 105020. https://doi.org/10.1016/j.jfca.2022.105020 . RD 4/ (2014). de 10 de enero, por el que se aprueba la norma de calidad para la carne, el jamón, la paleta y la caña de lomo ibérico. Realini, C. E., Duckett, S. K., & Windham, W. R. (2004). Effect of vitamin C addition to ground beef from grass-fed or grain-fed sources on color and lipid stability, and prediction of fatty acid composition by near-infrared reflectance analysis. Meat Science , 68 (1), 35–43. https://doi.org/10.1016/j.meatsci.2004.02.002 . Ripoll, G., Albertí, P., Panea, B., Olleta, J. L., & Sañudo, C. (2008). Near-infrared reflectance spectroscopy for predicting chemical, instrumental and sensory quality of beef. Meat Science , 80 (3), 697–702. https://doi.org/10.1016/j.meatsci.2008.03.009 . Rodríguez-Estévez, V., García, A., Peña, F., & Gómez, A. G. (2009). Foraging of Iberian fattening pigs grazing natural pasture in the dehesa. Livestock Science , 120 (1), 135–143. https://doi.org/10.1016/j.livsci.2008.05.006 . Rubio, B., Martínez, B., Sánchez, M. J., García-Cachán, M. D., Rovira, J., & Jaime, I. (2007). Study of the shelf life of a dry fermented sausage salchichon made from raw material enriched in monounsaturated and polyunsaturated fatty acids and stored under modified atmospheres. Meat Science , 76 (1), 128–137. https://doi.org/10.1016/j.meatsci.2006.10.021 . Ruiz-Carrascal, J., Ventanas, J., Cava, R., Andrés, A. I., & Garcı́a, C. (2000). Texture and appearance of dry cured ham as affected by fat content and fatty acid composition. Food Research International , 33 (2), 91–95. https://doi.org/10.1016/S0963-9969(99)00153-2 . Shahidi, F., & Zhong, Y. (2010). Lipid oxidation and improving the oxidative stability. Chemical Society Reviews , 39 (11), 4067–4079. https://doi.org/10.1039/B922183M . Sheard, P. R., Enser, M., Wood, J. D., Nute, G. R., Gill, B. P., & Richardson, R. I. (2000). Shelf life and quality of pork and pork products with raised n-3 PUFA. Meat Science , 55 (2), 213–221. https://doi.org/10.1016/S0309-1740(99)00145-X . Shenk, J. S., & Westerhaus, M. O. (1996). Calibration the ISI way. Near Infrared Spectroscopy: The Future Waves , 198–202. Tejeda, J. F., Gandemer, G., Antequera, T., Viau, M., & Garcı́a, C. (2002). Lipid traits of muscles as related to genotype and fattening diet in Iberian pigs: Total intramuscular lipids and triacylglycerols. Meat Science , 60 (4), 357–363. https://doi.org/10.1016/S0309-1740(01)00143-7 . Tejerina, D., García Torres, S., Cabeza de Vaca, M., Ortiz, A., & Romero, M. P. (2018). Evaluation of near-infrared spectroscopy (NIRS) for the quality control of packaged cured ham-sliced from Iberian pigs. Archivos de Zootecnia , 1 , 231–234. https://doi.org/10.21071/az.v67iSupplement.3611 . Toldrá, F., & Aristoy, M. C. (2004). Dry-cured ham. FOOD SCIENCE AND TECHNOLOGY-NEW YORK-MARCEL DEKKER- , 369–384. Toldrá, F., & Flores, M. (1998). The Role of Muscle Proteases and Lipases in Flavor Development During the Processing of Dry-Cured Ham. Critical Reviews in Food Science and Nutrition , 38 (4), 331–352. https://doi.org/10.1080/10408699891274237 . Torkashvand, A. M., Ahmadi, A., & Nikravesh, N. L. (2017). Prediction of kiwifruit firmness using fruit mineral nutrient concentration by artificial neural network (ANN) and multiple linear regressions (MLR). Journal of Integrative Agriculture , 16 (7), 1634–1644. https://doi.org/10.1016/S2095-3119(16)61546-0 . Vásquez, N., Magán, C., Oblitas, J., Chuquizuta, T., Avila-George, H., & Castro, W. (2018). Comparison between artificial neural network and partial least squares regression models for hardness modeling during the ripening process of Swiss-type cheese using spectral profiles. Journal of Food Engineering , 219 , 8–15. https://doi.org/10.1016/j.jfoodeng.2017.09.008 . Xie, L. H., Tang, S. Q., Wei, X. J., Sheng, Z. H., Shao, G. N., Jiao, G. A., Hu, S. K., Wang-Lin, & Hu, P. S. (2022). Simultaneous determination of apparent amylose, amylose and amylopectin content and classification of waxy rice using near-infrared spectroscopy (NIRS). Food Chemistry , 388 , 132944. https://doi.org/10.1016/j.foodchem.2022.132944 . Zamora-Rojas, E., Garrido-Varo, A., De Pedro-Sanz, E., Guerrero-Ginel, J. E., & Pérez-Marín, D. (2011). Monitoring NIRS calibrations for use in routine meat analysis as part of Iberian pig-breeding programs. Food Chemistry , 129 (4), 1889–1897. https://doi.org/10.1016/j.foodchem.2011.05.139 . Zhou, L. J., Wu, H., Li, J. T., Wang, Z. Y., & Zhang, L. Y. (2012). Determination of fatty acids in broiler breast meat by near-infrared reflectance spectroscopy. Meat Science , 90 (3), 658–664. https://doi.org/10.1016/j.meatsci.2011.10.010 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 29 May, 2024 Reviews received at journal 27 May, 2024 Reviewers agreed at journal 23 May, 2024 Reviews received at journal 07 May, 2024 Reviewers agreed at journal 03 May, 2024 Reviewers agreed at journal 18 Apr, 2024 Reviewers invited by journal 16 Apr, 2024 Editor assigned by journal 15 Apr, 2024 Submission checks completed at journal 14 Apr, 2024 First submitted to journal 09 Apr, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4241621","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":292110550,"identity":"4640e7ba-3759-4324-b7d9-f166941de5fc","order_by":0,"name":"Miriam Hernández-Jiménez","email":"","orcid":"","institution":"University of Salamanca","correspondingAuthor":false,"prefix":"","firstName":"Miriam","middleName":"","lastName":"Hernández-Jiménez","suffix":""},{"id":292110551,"identity":"6bac0897-d41c-40df-af98-1272f37172b7","order_by":1,"name":"Isabel Revilla","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvElEQVRIiWNgGAWjYBACNhCRUGDBwA9iPCBei4EEg2QDiEG8XUAtBgeI1cInffiYxAMDCXnjG+kPPyQw2NgTdhhfWpoE0GGG227kGEskMKQlNhDUwsNjbADUwgjUwgDUcpiw29h4+D+DtNhvnpH++EcCw38iHMbDw/gAqCVxg0SCGdCWA4xEOIzNEKQlecaZN2YWCQbJhP0i38P84OCPChvb/vb0xzc+VNgRdhgaMCBVwygYBaNgFIwCrAAAHy0zqA0H7WUAAAAASUVORK5CYII=","orcid":"","institution":"University of Salamanca","correspondingAuthor":true,"prefix":"","firstName":"Isabel","middleName":"","lastName":"Revilla","suffix":""},{"id":292110552,"identity":"d0506ad0-eb65-4c31-a336-0fdf069b4ea4","order_by":2,"name":"Pedro Hernández-Ramos","email":"","orcid":"","institution":"University of Salamanca","correspondingAuthor":false,"prefix":"","firstName":"Pedro","middleName":"","lastName":"Hernández-Ramos","suffix":""},{"id":292110553,"identity":"b349b134-f64c-4ff7-882f-a22cd7a6130f","order_by":3,"name":"Ana María Vivar-Quintana","email":"","orcid":"","institution":"University of Salamanca","correspondingAuthor":false,"prefix":"","firstName":"Ana","middleName":"María","lastName":"Vivar-Quintana","suffix":""}],"badges":[],"createdAt":"2024-04-09 11:46:48","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4241621/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4241621/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":55009630,"identity":"962250a4-7578-4445-9ffe-c2c1931ffdab","added_by":"auto","created_at":"2024-04-19 19:14:24","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":484770,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Raw spectra of the 148 samples and (b) mean NIR spectra after pre-treatment with the first derivative of the (\u003cstrong\u003e-\u003c/strong\u003e) shoulder 8 months, (\u003cstrong\u003e-\u003c/strong\u003e) ham 20 months, (\u003cstrong\u003e-\u003c/strong\u003e) shoulder 25 months, (\u003cstrong\u003e-\u003c/strong\u003e) ham 36 months\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4241621/v1/c7b21de14e1777e60962d71c.jpeg"},{"id":55009628,"identity":"6db02cbd-5e86-458b-973d-fe0c922b41aa","added_by":"auto","created_at":"2024-04-19 19:14:24","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":98291,"visible":true,"origin":"","legend":"\u003cp\u003eLinear regression plot of measured values versus predicted values by MPLS within the calibration set in samples of Iberian hams and Iberian shoulder\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4241621/v1/12c887b4c1de0a08eed94a64.png"},{"id":55010434,"identity":"6cd886fb-8de3-46c5-a43c-dc410770c16a","added_by":"auto","created_at":"2024-04-19 19:22:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":103640,"visible":true,"origin":"","legend":"\u003cp\u003eLinear regression plot of measured values versus predicted values by ANNs in samples of Iberian hams and Iberian shoulders\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4241621/v1/a58bfa3d18d6f706966e6d49.png"},{"id":55009627,"identity":"18a1985d-ba42-4f1d-aabf-5f1cb062b05e","added_by":"auto","created_at":"2024-04-19 19:14:24","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":98870,"visible":true,"origin":"","legend":"\u003cp\u003eThe distribution of regression coefficients from 1,100 to 2,000 nm for (a) palmitic (C16:0), (b) oleic (C18:1), and (c) linoleic (C18:2 n6) acids in Iberian ham and shoulder samples. The right axis corresponds to the MPLS method and the left axis corresponds to the ANN method\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4241621/v1/89e17a4dde4d463f906381e3.png"},{"id":55010749,"identity":"94adcae6-b608-4349-8297-aa95e9b0d256","added_by":"auto","created_at":"2024-04-19 19:30:25","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":729751,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4241621/v1/f7bec713-24f0-4ea9-a82b-e65e2b54b56d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003ePrediction of the Fatty Acid Profiles of Iberian Pig Products by Near Infrared Spectroscopy: A Comparison Between Multiple Regression Tools and Artificial Neural Networks\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eIberian ham is one of Spain's most renowned products (Toldr\u0026aacute; \u0026amp; Aristoy, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Ham and all other products made from the Iberian pig are characterized by their high percentage of fat, which is a key factor in its influence on technological quality and its impact on the sensory quality of the final products (Gandemer, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). This is the reason why so much research has been carried out to define the production factors affecting the intramuscular fat content and the composition of this fat. These factors include the breed, age, and live weight of the animal, but diet composition in the pre-slaughter period is considered to be the most influential (Carrapiso, Bonilla \u0026amp; Garc\u0026iacute;a, 2003). The higher quality of cured products from the Iberian breed is usually correlated with the traditional production systems which consist of feeding these Iberian pigs during the fattening period (\u003cem\u003emontanera\u003c/em\u003e) exclusively on natural resources found in a traditional landscape known as the \u003cem\u003edehesa\u003c/em\u003e, i.e. pastures and mainly acorns (Rodr\u0026iacute;guez-Est\u0026eacute;vez et al., \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eCurrent Spanish legislation classifies Iberian pig products (ham, shoulder, and loin) according to four categories which are Bellota 100% for products of the highest quality followed by Bellota Ib\u0026eacute;rico, Ib\u0026eacute;rico de cebo de campo, and Ib\u0026eacute;rico de cebo (RD 4/2014). The first two categories refer to animals fed during the \u003cem\u003emontanera\u003c/em\u003e (grazing free-range on grass and acorns) but the 100% Ib\u0026eacute;rico come from purebred Iberian animals and the Bellota Ib\u0026eacute;rico from Iberian x Duroc crossbred animals. This system of \u003cem\u003emontanera\u003c/em\u003e allows the pork to acquire higher intramuscular fat with a characteristic high proportion of unsaturated fatty acids (Cava et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Gonz\u0026aacute;lez \u0026amp; Tejeda, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Ruiz-Carrascal, Ventanas, Cava, Andr\u0026eacute;s \u0026amp; Garc\u0026iacute;a, 2000; Tejeda, Gandemer, Antequera, Viau, \u0026amp; Garc\u0026iacute;a 2002) which contributes to the unique and desirable sensory properties of the product. However, the higher degree of fat unsaturation also increases lipid oxidation and thus decreases shelf life (Rubio et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Shahidi \u0026amp; Zhong, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Sheard et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2000\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOn the one hand, it is nowadays strongly recommendable to reduce the intake of saturated fats and increase the intake of polyunsaturated fats, especially omega-3 fats, to prevent cardiovascular disease (Mataix, Quiles \u0026amp; Rodr\u0026iacute;guez, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). Consumers have therefore shown an increasing interest in the fat composition of their diets and demand more extensive knowledge of the fatty acid profiles of the products they purchase. On the other hand, the Iberian pork industry is seeking quality control systems which allow it to certify the authenticity of its products at any point in the distribution chain and provide accurate product information, setting the price according to quality (Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Ascertaining the fatty acid profiles of products such as ham and shoulder could provide important information related to quality control, sensory attributes, shelf life, and nutritional properties (Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Tejerina, Garc\u0026iacute;a-Torres, Cabeza de Vaca, Ortiz \u0026amp; Romero-Fern\u0026aacute;ndez, 2018).\u003c/p\u003e \u003cp\u003eThe analysis of quality parameters in general and the lipid profile in particular involves sample destruction and is expensive and time-consuming, which is a problem for the meat industry (Tejerina et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Non-invasive technologies are therefore useful alternatives for quality analysis and product authentication. In this scenario, low-field 1H-NMR has been used to determine the lipid profile of Iberian ham; furthermore, classification according to current commercial categories was possible using chemometric tools (Pajuelo et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Ion Mobility Spectrometry coupled with Gas Chromatography (IMS-GC) and the use of non-destructive sampling methods has also been shown to be capable of classifying ham samples according to their commercial category by using volatilome (Mart\u0026iacute;n-G\u0026oacute;mez, Arroyo-Manzanares, Rodr\u0026iacute;guez-Est\u0026eacute;vez \u0026amp; Arce, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNear Infrared Spectroscopy (NIRS) is one of the most outstanding non-invasive technologies as it is a fast, non-destructive, and accurate tool for predicting the chemical composition of meat (Alomar, Gallo, Casta\u0026ntilde;eda \u0026amp; Fuchslocher, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). The main advantage of NIR spectroscopy is that it allows a multitude of analytical parameters to be determined in a few minutes, facilitating a more exhaustive control of a large number of individuals or pieces (Prieto, Pawludzyk, Dugan\u0026amp; Aalhus, 2017; Zamora-Rojas, Garrido-Varo, De Pedro-Sanz, Guerrero-Ginel \u0026amp; P\u0026eacute;rez-Mar\u0026iacute;n, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In the Iberian sector several research studies have been carried out using quantitative models which have shown a high potential for predicting the fatty acid and stable carbon isotope profile in subcutaneous fat of the carcass and/or live animal (Garc\u0026iacute;a-Olmo, Garrido-Varo, De Pedro, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Gonz\u0026aacute;lez-Mart\u0026iacute;n et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; P\u0026eacute;rez-Mar\u0026iacute;n et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Qualitative models have also allowed classification according to the diet, breed and management systems by using spectra recorded both in the live animal and in different areas of the carcass (Arce et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Hern\u0026aacute;ndez-Jim\u0026eacute;nez, Gonz\u0026aacute;lez-Mart\u0026iacute;n, Mart\u0026iacute;nez-Mart\u0026iacute;n, Revilla \u0026amp; Vivar-Quintana \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Horcada, Valera, Ju\u0026aacute;rez \u0026amp; Fern\u0026aacute;ndez-Caban\u0026aacute;s, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; P\u0026eacute;rez-Mar\u0026iacute;n, Fearn, Riccioli, De Pedro \u0026amp; Garrido, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe use of NIR spectroscopy in cured products has allowed the prediction of the main quality characteristics in a fast and non-destructive manner (Ortiz, Le\u0026oacute;n, Contador \u0026amp; Tejerina, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, studies of dry-cured ham are rather scarce and include the prediction of texture and color (Garc\u0026iacute;a, Andrada, Mu\u0026ntilde;oz \u0026amp; Bote, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), the sensory characteristics (Hern\u0026aacute;ndez-Ramos et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) or salt content, and the fatty acid oxidation and tocopherol content in sliced and packaged ham (Tejerina et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) while the investigation of the feasibility of fatty acid profile prediction has been addressed by only a few studies (Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Tejerina et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The determination of the fatty acid profile of the final product by NIRS would facilitate the characterization of the product in the slicing and packaging stage. Furthermore, it would allow its authentication by relating the spectra to the diet and finally it would be a very useful tool for determining the fat saturation index so as to correlate it with the shelf life of the product (Ansorena \u0026amp; Astiasar\u0026aacute;n, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Rubio et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Sheard et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2000\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOne of the drawbacks of NIRS spectroscopy is data processing (Jurinjak Tušek et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Several traditional methods are used such as Principal Component Analysis (PCA), Principal Component Regression (PCR), Canonical Correlation Analysis (CCA), Multiple Linear Regression (MLR), and in particular Partial Least Square Regression (PLSR) (Balabin \u0026amp; Safieva, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; P\u0026eacute;rez-Mar\u0026iacute;n, Garrido-Varo, De Pedro \u0026amp; Guerrero-Ginel, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). All of these are based on linear fit regressions. The nonlinearity of the data deriving from NIR spectroscopy is the main problem of these multivariate methods (Balabin, Safieva \u0026amp; Lomakina, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, Balabin, Safieva \u0026amp; Lomakina, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Therefore, the search for more complex algorithms with nonlinear and nonparametric regression has been resorted to; these algorithms include artificial neural networks (ANN) which have shown great potential (P\u0026eacute;rez-Mar\u0026iacute;n et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). ANNs are able to recognize and reproduce cause-effect relationships through a multitude of training in input-output systems (Kundu, Paul, Kumar \u0026amp; Mishra, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In the agri-food sector they have been used in combination with NIRS spectroscopy for the prediction of sensory parameters in raw-cured meat products (Hern\u0026aacute;ndez-Jim\u0026eacute;nez et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hern\u0026aacute;ndez-Ramos et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), the prediction of textural parameters in cheese (V\u0026aacute;squez et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and quality determination in fruits (Alhamdan, Fickak \u0026amp; Atia, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). No previous studies have been carried out with the aim of predicting fatty acids with neural networks in complex matrices such as meat, but studies of lipid profile determination in oils do exist (Rajković et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe aim of this study was to determine the potential of NIR spectroscopy for the characterization of the lipid profile of Iberian ham and shoulder samples at different curing times prior to their commercialization. Two different methods were evaluated for the development of predictive models, one based on linear regression (MPLS) and the other based on non-linear regression (ANN). Several training algorithms and network architectures were evaluated for this latter method prior to the development of the model.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Samples\u003c/h2\u003e \u003cp\u003eA total of 148 pieces (76 shoulders and 72 hams) from Iberian pigs slaughtered at 145\u0026thinsp;\u0026plusmn;\u0026thinsp;15 kg carcass weight were selected for analysis. All the animals were fattened during the \u003cem\u003emontanera\u003c/em\u003e, i.e. fed on grass and acorns of \u003cem\u003eQuercus ilex\u003c/em\u003e and \u003cem\u003eQuercus suber\u003c/em\u003e for 68\u0026ndash;120 days depending on the management system of the farm of origin. Half of the animals were of 100% Iberian breed and the other half of 50% Iberian x Duroc breed. The pieces were processed by the local industry (Carrasco Ib\u0026eacute;ricos, Guijuelo, Spain) according to traditional methods consisting of a salting stage in salt piles at low temperature (1\u0026ndash;4\u0026ordm;C) and high relative humidity (83\u0026ndash;85%) for 1 day per kg of weight of the piece. At the end of this stage the surface was brushed and washed to remove the surface salt; the pieces subsequently underwent the post-salting stage for 9 months at temperatures of 4\u0026ndash;5\u0026ordm;C and a relative humidity of 80%. They were then moved to drying rooms where the temperature was progressively increased from 4 to 27\u0026ordm;C while the relative humidity was decreased from 70\u0026ndash;50% for 4\u0026ndash;5 months. Finally, the hams and shoulders underwent a cellar stage with temperatures ranging from 17 to 25\u0026ordm;C and a relative humidity of 55\u0026ndash;60%. Throughout this process intermediate sampling was carried out on a total of 16 shoulders at 250 days (8 months) and a total of 12 hams at 585 days (20 months). In addition, 60 pieces of shoulder and 60 pieces of ham were sampled at the end of the cellar process with processing times of 768 days (25 months) and 1075 days (36 months) respectively.\u003c/p\u003e \u003cp\u003eFor the sampling of the pieces, the bone was removed and a slice 3 cm thick was cut in the central area of the piece along its entire transverse profile, including all the muscles. The cut was made perpendicular to the bone and above it at the same height in all the pieces. This type of sampling allows the analysis of physical and chemical properties in addition to the NIRS recording of the different muscles identified. To analyze the lipid profile, the \u003cem\u003eBiceps femoris\u003c/em\u003e muscles in the case of the hams and the \u003cem\u003eBrachiocephalic\u003c/em\u003e muscle in the case of the shoulders were trimmed and kept at temperatures of -32\u0026ordm;C until the time of analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Fatty acid analysis\u003c/h2\u003e \u003cp\u003eThe fatty acid profile was determined by using gas chromatography. Total lipid extraction was carried out by using the method described by Folch, Lees \u0026amp; Sloane Stanley (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1957\u003c/span\u003e), which combines mechanical extraction with organic solvent extraction in two different muscles, \u003cem\u003eBiceps femoris\u003c/em\u003e in the case of hams and \u003cem\u003eBrachiocephalicus\u003c/em\u003e in the case of shoulders. Subsequent methylation was performed on 0.1 grams of extracted fat by using the method described by Gonz\u0026aacute;lez-Mart\u0026iacute;n, Vicente Palacios, Revilla, Vivar-Quintana \u0026amp; Hern\u0026aacute;ndez-Hierro (2017). A GC 6890 N (Agilent Technologies, USA) chromatograph equipped with an automatic injector 7683 (Agilent Technologies) and a fused silica capillary column (100 m \u0026times; 0.25 mm; 0.20 \u0026micro;m silica (SP-2560, Supelco, Inc., Bellefonte, PA, USA) were used. The injector and detector were maintained at 250\u0026ordm;C. The column oven temperature was 150\u0026ordm;C and the temperature was increased 1\u0026ordm;C/min to 165\u0026ordm;C, then 0.20\u0026ordm;C/min to 167\u0026ordm;C, and finally increased 1.50\u0026ordm;C/min to 225\u0026ordm;C for 15 min. The carrier gas was helium at a flow rate of 1ml/min. The injection volume was 1 \u0026micro;L in splitless mode. The different fatty acids were identified by their retention times using a mixture of fatty acid standards (47885-U Supelco, Sigma-Aldrich, Germany). Fatty acid contents were calculated by using the peak areas of the chromatograms and expressed as g per 100 g of the total fatty acid methyl esters.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Near infrared Spectroscopy (NIRS) spectra register\u003c/h2\u003e \u003cp\u003eThe NIRS spectra of the samples obtained using a Foss NIRSystem 5000 (Hillerod, Denmark) were recorded using a fiber-optic probe (1.5 m 210/210, Ref. n\u0026deg; R6539-A) coupled with a 5 cm x 5 cm window quartz. The spectra of the sample were recorded in the 1100\u0026ndash;2000 nm range at intervals (2 nm), which means that a total of 451 data were obtained for each sample and 32 scans were performed for each recording. The window was applied directly to the surface of the ham or shoulder slice. In the ham, the recording was performed on the \u003cem\u003eBiceps femoris\u003c/em\u003e. In the shoulder, the spectral recording was performed on the \u003cem\u003eBrachiocephalicus\u003c/em\u003e muscle. NIRS measurements were taken in diffuse reflectance and as a first step the spectra were converted from reflectance to absorbance as log 1/R. The three absorbance spectra of each sample were visually examined for consistency and then averaged (the outlier spectra were removed). The Win ISI 4.10 software package was used for spectra collection and data handling.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Calibration\u003c/h2\u003e \u003cp\u003eTo predict the lipid profile jointly in both types of muscles (\u003cem\u003eBiceps femoris\u003c/em\u003e and \u003cem\u003eBrachiocephalicus\u003c/em\u003e), calibration was carried out by using two different methods in order to compare them. The first was the modified partial least squares regression method (MPLS) and the second was artificial neural networks (ANNs). Of the total fatty acids determined in ham and shoulder muscles at different curing times, we selected those that could be quantified in at least 90% of the samples analyzed.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2.4.1 MPLS Calibration\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe initial data for the calibrations were the 451 spectral data recorded in the near infrared region (1100\u0026ndash;2000, 2 nm) together with the quantified percentage of each fatty acid (C14:0, C16:0, C18:0, C16:1, C18:1, C18:1 n7, C22:1 n9, C18:2 n6, C18:3 n3, and C20:2 n6) and their corresponding sums (SFA, MUFA, PUFA, n3 and n6) from each of the ham and shoulder samples (148 samples). These data sets were used as input and output data respectively. Possible anomalous or outlier spectra was detected by using principal component analysis (PCA) which also reduced the number of variables. The Mahalanobis distance indicates how different a spectrum is from the average spectrum of the set (De Maesschalck, Jouan-Rimbaud \u0026amp; Massart, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). When H took values of H\u0026thinsp;\u0026gt;\u0026thinsp;3.5 the spectrum was considered anomalous and eliminated from the population. The criterion T\u0026thinsp;\u0026gt;\u0026thinsp;2.5 was also applied in order to eliminate from the calibration set those samples with differences between the laboratory-determined fatty acid percentage and the NIRS predicted value. Because NIR spectra are affected by the physical characteristics of the sample, i.e. light scattering and wavelength variations, scattering treatments were applied including multiplicative scatter correction (MSC), standard normal variable (SNV), DeTrend (DT) or SNV-DT, and both first and second order derivatives (Marini, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Norris, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). The regression method to obtain the lipid profile prediction equations using the NIR spectra as independent variables was the MPLS (modified partial least squares) method. MPLS is an improved form of PLS developed by FOSS and obtains better results in handling small spectral variability (Xie et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The total spectra were divided into two subsets: a calibration set and a validation set. During the development of the different equations, a cross-validation process was carried out in order to select the optimal number of factors and avoid overfitting (P\u0026eacute;rez-Mar\u0026iacute;n et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) and the equation with the best predictive capacity was selected according to the following criteria: the highest determination coefficient in the calibration (RSQ or R2) and the lowest standard error calibration (SEC) and standard error cross-validation (SECV). In addition, the residual prediction deviation (RPD), which can be defined as the ratio of the standard deviation (SD) of the reference values over the calibration standard error (RPD\u0026thinsp;=\u0026thinsp;SD/SEC), was calculated and used to evaluate the practical usefulness of predictive models (Hruschaka, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2001\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFinally, the method was tested by applying the calibration equations obtained to the validation set of samples, which allows the predicted values to be compared with the actual data. The standard error of prediction corrected (SEPC) and the root mean square error (RMSE) statistics were obtained; these are indicative of the prediction. The predicted values were compared with the reference values using Student's t-test for paired values. Calibration and validation of the models was performed using the WinISI Version 4.10 (Infrasoft International, State Collee, PA, USA).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.4.2 Artificial Neural Network Calibration\u003c/h2\u003e \u003cp\u003eThe neural network models were built with MatLab (R2023a, MathWorks Inc.) as a multilayer feed-forward network. As in the case of the MPLS calibration, the input data were the 451 spectral data recorded in the NIR region and the output data were the quantified percentages of each fatty acid and their corresponding sums of the 148 samples analyzed. The ANN architecture can be described as having an input layer, a hidden layer with a variable number of neurons, and a single neuron in the output layer where the value to be predicted has been obtained. The transfer functions implemented were the sigmoid hyperbolic tangent function for the neurons in the hidden layer and the rectified linear unit (RLU) function for the neuron in the output layer respectively. In total, hundreds of architectures were analyzed with a variable number of neurons in the hidden layer (between 1 and 25) for each parameter, each of which was trained 300 times starting from initial weights randomly assigned from a known seed value, which allows the results to be reproduced (Pillonel et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2005\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe cost function reduction (MSE) was tested with the Scaled Conjugate Gradient, Descendent Gradient, Descendent Gradient with Variable Learning Rate, Descendent Gradient with Momentum, Bayesian Regularization, Quasi-Newton and Levenberg-Marquardt algorithms. The analytical data pairs, i.e. the NIRS data and fatty acid percentages, were randomly divided into three sets to train the network. The first set was a training set (70% of the samples), the second a validation set (15% of the samples), and the third a test set to measure the goodness of the network (15% of the samples). The ANN architectures were then optimized for each of the fatty acids and their summations. The predictive ability of the models was determined by a higher RSQ and a lower RMSE.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Results and discussion","content":"\u003cp\u003e3.1 Fatty acid composition\u003c/p\u003e\n\u003cp\u003eThe lipid composition of the pieces analyzed both at an intermediate curing point and in the final product is presented in Table 1. The results show a wide variability, which is of interest when pursuing the development of calibration equations. A total of 31 fatty acids were identified and quantified. Of these, the most abundant were monounsaturated fatty acids (MUFA) with percentages between 59.59 and 62.06% with a significantly higher content in the group of intermediate cured hams (20 months). The sum of saturated fatty acids (SFA) presented values of between 31.42 and 32.06 with no significant differences between the four products, while the values of polyunsaturated fatty acids (PUFA) varied between 6.52 and 8.76% with higher values found in the shoulders than in the hams and higher values in the final products than in the intermediate samples. The mean values of the different summations were similar to those obtained by other studies for Iberian ham from animals fattened with the same production system (\u003cem\u003emontanera\u003c/em\u003e) and the same breeds (100% and 50% Iberian)\u0026nbsp;(Pajuelo et al., 2022). For the sum of the n3, significant differences were found in the four groups analyzed with significantly higher values in the products at the end of the curing period. In the case of the n6, the shoulders showed significantly higher values than the hams in both samples.\u003c/p\u003e\n\u003cp\u003eTable 1. Mean values (\u0026plusmn; standard deviation) of the fatty acids analysed in shoulder and ham at the intermediate point of curing and in the final product, expressed as % by weight of total fatty acids.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.567164179104477%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"20.149253731343283%\" colspan=\"4\" rowspan=\"3\"\u003e\n \u003cp\u003eShoulder\u003c/p\u003e\n \u003cp\u003e8 months\u003c/p\u003e\n \u003cp\u003eN=16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.402985074626866%\" colspan=\"4\"\u003e\n \u003cp\u003eHam\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.402985074626866%\" colspan=\"4\"\u003e\n \u003cp\u003eShoulder\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.96268656716418%\" colspan=\"4\"\u003e\n \u003cp\u003eHam\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.514925373134329%\" colspan=\"3\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"14.485981308411215%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.299065420560748%\" colspan=\"4\"\u003e\n \u003cp\u003e20 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.299065420560748%\" colspan=\"4\"\u003e\n \u003cp\u003e25 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25%\" colspan=\"4\"\u003e\n \u003cp\u003e36 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.91588785046729%\" colspan=\"3\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"14.485981308411215%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.299065420560748%\" colspan=\"4\"\u003e\n \u003cp\u003eN=12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.299065420560748%\" colspan=\"4\"\u003e\n \u003cp\u003eN=60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25%\" colspan=\"4\"\u003e\n \u003cp\u003eN=60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.91588785046729%\" colspan=\"3\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.567164179104477%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"9.701492537313433%\" colspan=\"2\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.634328358208956%\" colspan=\"2\"\u003e\n \u003cp\u003eSd\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.76865671641791%\" colspan=\"2\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.634328358208956%\" colspan=\"2\"\u003e\n \u003cp\u003eSd\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.76865671641791%\" colspan=\"2\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.634328358208956%\" colspan=\"2\"\u003e\n \u003cp\u003eSd\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.76865671641791%\" colspan=\"2\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.007462686567164%\" colspan=\"2\"\u003e\n \u003cp\u003eSd\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.955223880597014%\" colspan=\"2\"\u003e\n \u003cp\u003eP-valor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"0.5597014925373134%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC12:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,099\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC14:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e1,41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,36a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,10a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,65b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,15a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC15:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,764\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC16:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e21,67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e20,91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e22,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e21,91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,130\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC17:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,439\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e8,18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,61ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e8,72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,94b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e7,33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,31a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e8,23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,97ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,09b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC21:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,05c\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC22:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,050\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC23:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,85b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,33ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,006\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC24:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,05b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003eSFA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e32,06\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;2,55\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e31,42\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,82\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e31,67\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;2,24\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e32,06\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;2,10\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e0,667\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC14:1 n5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,806\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026sum; C16:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e4,81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,00ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,62a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e5,24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,37b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,53\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,58ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC17:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,275\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:1 n9t\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,435\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e50,23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;2,75a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e52,92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,33b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e49,19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;2,69a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e50,68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,61a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:1 n7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e4,48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,28ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,51a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,45ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,48b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:1 n9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,265\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC22:1 n9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,10c\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,14bc\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC24:1 n9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003eMUFA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e60,30\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;2,33a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e62,06\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,95b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e59,59\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;2,09a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e60,47\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,53a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e0,000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:2 n6t\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,07b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,10b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e5,75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,82b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e4,81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;1,23a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e6,80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,88c\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e5,38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,90ab\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:3 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,12b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC18:3 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e1,27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,26b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,07a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,15c\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e1,33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,12b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,13b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:3 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,323\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:3 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,05b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:4 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,679\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC22:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,143\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC20:5 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,512\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003eC22:6 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,00a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,01a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,08a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e0,06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026plusmn;0,02b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e0,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003ePUFA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e7,64\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,23b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e6,52\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,34a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e8,76\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,04c\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e7,45\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,10b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e0,000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003en3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e1,28\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,26b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e1,02\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,07a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e1,48\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,19c\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e1,64\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,15d\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e0,000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.962616822429906%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003en6\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.719626168224298%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e6,68\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,26b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e5,41\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;1,22a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e7,28\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.654205607476635%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,96b\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.785046728971963%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e5,81\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.02803738317757%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026plusmn;0,97a\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.97196261682243%\" colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003e0,000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"11.481481481481481%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"9.074074074074074%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"10%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"8.333333333333334%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"10%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"8.333333333333334%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"10%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"8.333333333333334%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"10.555555555555555%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"8.333333333333334%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.5555555555555556%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe number of fatty acids quantified differed depending on the type of muscle and stage of maturation; it was found that this number was higher in samples from the \u003cem\u003eBiceps femoris\u0026nbsp;\u003c/em\u003e(ham). In all cases it was possible to quantify the characteristic and main fatty acids of Iberian pig fat. Among them are oleic acid (C18:1) with values of between 49.19 and 52.91%, palmitic acid (C16:0) with values of between 20.91 and 22.01%, stearic acid (C18:0) with values of between 7.33 and 8.72%, and linoleic acid (C18:2 n6) with values of between 4.81 and 6.80%. The results obtained coincide with those reported for ham from animals which have fed extensively on acorns and grass (Mart\u0026iacute;n-C\u0026aacute;ceres, 1996). In addition to these major fatty acids, it was possible to quantify other fatty acids such as palmitoleic (C16:1), trans-vaccenic (C18:1 n7), myristic (C14:0) and \u0026alpha;-linolenic (C18:3 n3) acids which presented lower values between 1.01 and 5.24%.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWith regard to the effect of the curing time, significant differences were observed in 12 out of the 31 fatty acids quantified between the intermediate sampling (20 months) and the final sampling of the hams (36 months). In the case of the shoulders sampled at 8 and 25 months, these differences were only significant in 4 out of the 31 fatty acids quantified. This fact is due to the processes of lipolysis and oxidation during this processing period, which in turn are influenced by different factors such as the length of the curing period, the level and manner of salting, and the environmental conditions during maturation\u0026nbsp;(Buscailhon \u0026amp; Monin, 1994; Toldr\u0026aacute; \u0026amp; Flores, 1998). Furthermore, they may also depend on the type of muscle fiber, as this influences changes in the oxidative stability of the meat, or on lipid oxidation phenomena during storage and processing\u0026nbsp;(A. I. Andr\u0026eacute;s et al., 2001)\u0026nbsp;which are related to the presence of prooxidant and antioxidant factors\u0026nbsp;(Aristoy \u0026amp; Toldr\u0026aacute;, 1998; Henckel, Oksbjerg, Erlandsen, Barton-Gade \u0026amp; Bejerholm, 1997).\u003c/p\u003e\n\u003cp\u003eAs for the effect of maturation on the different fatty acids, this was similar in both shoulders and hams. Myristic (C14:0), palmitic (C16:0), palmitoleic (C16:1), linoleic (C18:2 n6), and \u0026alpha;-linolenic (C18:3 n3) fatty acids and the sum of SFA and PUFA n3 showed an increase during maturation in both hams and shoulders and these differences were significant for both C14:0 and C18:2 n6 in the case of shoulders. In contrast, a decrease was observed in the percentage of other fatty acids such as C18:0, C20:0, C21:0, C23:0, C18:1, C22:1 n9, C18:2 n6t, and C18:3 n6 and also in the sum of MUFA. In this case the differences were generally significant for hams except for C20:0 for which shoulders showed a significant decrease owing to maturation. In general, it was therefore observed that maturation involved a decrease in MUFA and a significant increase in PUFA.\u003c/p\u003e\n\u003cp\u003eThe same behavior observed in this study for C16:0, C16:1, and C18:0 was also reported by Cava, Est\u0026eacute;vez, Morcuende \u0026amp; Antequera, (2003) in the \u003cem\u003eSemimembranosus\u0026nbsp;\u003c/em\u003emuscle. Antequera et al. (1992) also observed a decrease in certain unsaturated triglycerides during the curing process of ham mainly concerning lipolysis and direct oxidation phenomena.\u003c/p\u003e\n\u003cp\u003e3.2 Spectral characteristics\u003c/p\u003e\n\u003cp\u003eThe mean spectra of each of the 148 samples in the wavelength range of 1100 to 2000 nm were similar in type. Although a wide dispersion of the spectra could be observed (Figure 1a), the averages calculated according to product type and maturation time were very close to each other with the exception of the average for 36-month-old ham, which had higher absorbance values. This fact is related to physical factors owing to fibrillar changes caused by proteolysis processes, which are more evident in pieces with a longer curing time. The same behavior has been observed in fresh beef and lamb samples\u0026nbsp;(S. Andr\u0026eacute;s et al., 2007; Prieto, Andr\u0026eacute;s, Gir\u0026aacute;ldez, Mantec\u0026oacute;n \u0026amp; Lav\u0026iacute;n, 2008; Ripoll, Albert\u0026iacute;, Panea, Olleta \u0026amp; Sa\u0026ntilde;udo, 2008).\u003c/p\u003e\n\u003cp\u003eFor a better interpretation of the absorption zones, Figure 1b shows the pre-treated spectrum with an initial derivative (1,1,1,1), in which it can be seen that no anomalous spectra or spectral noise appeared throughout the measurement range\u0026nbsp;(Hruschaka, 2001). This representation of the spectrum allows the identification of the absorption maxima and the bands in which the variance is more marked\u0026nbsp;(Zhou, Wu, Li, Wang \u0026amp; Zhang, 2012). These maxima were reached in the second and first overtone bands and are associated with fat owing to the presence of C-H bonds at the 1200, 1730, and 1770 nm wavelengths\u0026nbsp;(Barlocco, Vadell, Ballesteros, Galietta \u0026amp; Cozzolino, 2006; Leroy et al., 2004; Prieto et al., 2008). Other absorption maxima were located in the first and second overtone at 1458 and 1824 nm which correspond to O-H and C-O bonds.\u003c/p\u003e\n\u003cp\u003eThe maximum dispersion was located in the bands related to water absorption (1140-1150 nm; 1300-1450 nm), to fats in the second overtone (1165-1200 nm; 1210-1270 nm), to the absorption of fat, water, and protein in the combination bands, (1300-1450 nm) and to the absorption of fats and aromatic compounds in the first overtone (1640-1725 nm; 1734 nm; 1770 nm). The 1680 nm band which indicates the presence of double bonds corresponding to unsaturated fatty acids is of special interest (Garrido-Varo, Carrete \u0026amp; Fern\u0026aacute;ndez-Caban\u0026aacute;s, 1998).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e3.3 Fatty acid calibration equations\u003c/p\u003e\n\u003cp\u003eIn order to develop the equations, the fatty acids which could be quantified in at least 90% of the samples were selected. According to this criterion, a total of 15 constituents were selected corresponding to 10 individual fatty acids and the different summations according to the number of unsaturation and their position, i.e. the summations of saturated (SFA), monounsaturated (MUFA), polyunsaturated (PUFA), and the summations of PUFA n3 and PUFA n6 fatty acids.\u003c/p\u003e\n\u003cp\u003e3.3.1 Calibration equations by the MPLS method\u003c/p\u003e\n\u003cp\u003eAs described above, for the development of equations using MPLS a PCA was performed to eliminate possible spectral outliers according to the H \u0026gt; 3.5 criterion. The total set of samples was then divided into a calibration set and an external validation set consisting of 80% (118 samples) and 20% (30 samples) respectively. Potential outliers by the reference method were also removed according to the T \u0026gt; 2.5 criterion. For each constituent a variable number was eliminated and equations which eliminated more than 10% of the samples were not considered valid; this maximum percentage of outlier samples was similar to that used in other studies\u0026nbsp;(Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., 2011). Based on this criterion, equations could be developed for 13 of the constituents, i.e. 8 fatty acids and the different summations. For each constituent the model was optimized by applying different mathematical and scatter treatments (None, SNV and DT) to obtain the maximum information provided by the chemical signals. It should be noted that the best calibration results were obtained with the mathematical pre-treatments applying a first and second derivative, which is in agreement with other studies related to the prediction of fatty acids in fats and products deriving from Iberian pork\u0026nbsp;(Fern\u0026aacute;ndez-Caban\u0026aacute;s, Garrido-Varo, Garc\u0026iacute;a-Olmo, De Pedro \u0026amp; Dardenne, 2007, Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., 2011).\u003c/p\u003e\n\u003cp\u003eDuring the development of the equations the model was evaluated by cross-validation. In this case the total number of samples was divided into four sets, using three for calibration and one for prediction; the process was repeated as many times as the number of sets available so that all the sets formed part of the calibration and prediction. This process allowed the validation of the model by checking the predictive capacity by means of the SECV statistic.\u003c/p\u003e\n\u003cp\u003eOf the total equations, 8 presented RSQ values of \u0026gt; 0.5 and RPD values of between 1.41 and 1.93 and corresponded to those fatty acids or summations which are present in the samples at levels higher than 1% (1.33-60.02%) and with a higher variability coefficient (4.37-22.30%) (Table 2). These constituents were the sum of the total PUFA, n3, and n6 and the fatty acids C14:0, C18:0, C18:1, C18:2 n6 and C18:3 n3. For the fatty acids with mean values of \u0026lt; 1% no equations could be developed. The prediction models obtained with RSQ values of between 0.30 and 0.69 can be used to separate samples with higher and lower analytical values, in this case the fatty acids C14:0, C16:0, C18:0, C16:1, C18:1, C18:1 n7, C18:2 n6, SFA, MUFA, and n6. RSQ values above 0.70 would indicate good predictive ability as is the case for C18:3 n3, PUFA, and n3 (Shenk \u0026amp; Westerhaus, 1996). It can also be observed that the determination coefficients of the summations are influenced by the calibration statistics of their main individual fatty acids, which has also been reported in other studies (Fern\u0026aacute;ndez-Caban\u0026aacute;s, Polvillo, Rodr\u0026iacute;guez-Acu\u0026ntilde;a, Botella \u0026amp; Horcada, 2011). The best calibration results were thus obtained for polyunsaturated fatty acids and therefore better results are obtained in summates of this group (Table 2). The SECV values obtained in most cases were higher than those determined in other studies predicting fatty acids in meat and meat products made from pork (Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., 2011; Gonz\u0026aacute;lez-Martı́n, Gonz\u0026aacute;lez-P\u0026eacute;rez, Alvarez-Garc\u0026iacute;a \u0026amp; Gonz\u0026aacute;lez-Cabrera, 2005; Tejerina et al., 2018).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2. Descriptors of NIR calibration and results of cross-validation for the Iberian ham and shoulder fatty acid profile by MPLS\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"747\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eFatty acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eMathematic treatment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003eTerms\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003eRange\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003eEst. Min\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003eEst. Max\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003eSEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003eSECV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003eRSQ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003eCV %\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003eRPD\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC14:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eSNV 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e106\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e1.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e0.84-2.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e2.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e22.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC16:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eDT 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e21.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e17.20-25.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e17.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e26.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.53\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e1.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e6.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.23\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC18:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eSNV+DT 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e105\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e7.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e5.50-10.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e4.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e10.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e13.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.75\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eSFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eDT 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e113\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e31.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e27.01-37.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e25.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e38.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e2.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e2.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e6.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC16:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eDT 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e109\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e4.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e2.89-6.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e2.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e6.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e15.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.20\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC18:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eDT 1,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e105\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e50.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e44.85-55.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e43.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e56.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e2.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e1.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e4.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC18:1 n7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eSNV 2,10,10,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e4.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e3.39-5.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e3.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e5.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e10.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.33\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eMUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eDT 0,0,1,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e60.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e54.68-63.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e54.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e65.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e3.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC18:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eNone 1,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e110\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e5.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e3.81-8.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e2.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e9.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e18.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.71\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003eC18:3 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eSNV+DT 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e105\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e1.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e0.94-1.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e1.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e12.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003ePUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eNone 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e8.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e5.72-10.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e4.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e11.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e1.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e15.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003en3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eSNV 2,4,4,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e1.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e1.2-2.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e1.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e2.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e12.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.85\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"8.981233243967829%\"\u003e\n \u003cp\u003en6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.75603217158177%\"\u003e\n \u003cp\u003eNone 2,10,10,1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.289544235924933%\"\u003e\n \u003cp\u003e112\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.166219839142091%\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.898123324396783%\"\u003e\n \u003cp\u003e6.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.198391420911529%\"\u003e\n \u003cp\u003e3.32-9.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.506702412868632%\"\u003e\n \u003cp\u003e2.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.908847184986596%\"\u003e\n \u003cp\u003e9.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"5.7640750670241285%\"\u003e\n \u003cp\u003e18.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"4.6916890080428955%\"\u003e\n \u003cp\u003e1.41\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eEst. Min= Estimated minimum; Est. Max= Estimated maximum; SD= Standard Deviation; SEC= Standard Error of Calibration. SECV: Standard Error of Cross-validation; RSQ= Coefficient of determination; CV= Coefficient of variation; RPD= Ratio of Performance to Deviation (SD/SEC)\u003c/p\u003e\n\u003cp\u003eThe MPLS calibration allows the determination of which wavelengths are the most important in the generation of the predictive equations because they are multiplied by a beta coefficient (\u0026beta;) with a higher value\u0026nbsp;(Lucas, Andueza, Ferlay \u0026amp; Mart\u0026iacute;n, 2008; Mart\u0026iacute;nez-Mart\u0026iacute;n, Hren\u0026aacute;ndez-Jim\u0026eacute;nez, Revilla y Vivar-Quintana, 2023). For each of the fatty acids therefore, the wavelengths with the highest \u0026beta; values were selected and correlated with the bonds and functional groups (information provided by WinISI software) which make the greatest contribution to absorbance at that wavelength (Table 3). In general, the wavelengths with the highest contribution are associated with molecules containing C-H, N-H, O-H, C=O, and S-H bonds with molecular vibrational modes mainly in the first overtone and to a lesser extent in the combination and second overtone bands.\u003c/p\u003e\n\u003cp\u003eTable 3. Spectral wavelengths and chemical groups with high \u0026beta; coefficients in the development of calibration equations\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"586\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eFatty acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003eWavelength (nm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eChemical bonds\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eChemical Structure and Funcional Groups\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC14:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1414\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eO-H, C-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eROH-H2O, CH, CH Oil\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1442\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eCH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1460\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmides\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1480\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmides\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1922\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC=O, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC16:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1470\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmides\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1480\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmines\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1922\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC=O, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC18:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1454\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1472\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmides\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1496\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eAmides\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1540\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eSFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1506\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eProtein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1518\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eUrea\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1542\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC16:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1480\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eGlucose\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1520\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eUrea\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1574\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC18:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1410\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eO-H, C-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater, C-H Oil\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1722\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H Oil\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1746\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eS-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eS-H\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1800\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1918\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC=O\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC18:1 n7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1364\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CH3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1534\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-RNH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1622\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e=CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eMUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1494\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H, O-H, N-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONHR, ARNH2, Urea\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1902\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC=O\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CO2H\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1920\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC=O\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC18:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1166\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eHC=CH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1442\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC-H, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CH2, Aromatic compounds\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1510\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eProtein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1614\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eO-H, C-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eROH-Water, -CH, Aromatic compounds\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1938\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003eWater\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003eC18:3 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1454\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-STARCH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1568\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1620\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e=CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1816\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-Cellulose\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003ePUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1444\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC-H, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CH2, Aromatic compounds\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1460\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eO-H, N-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1496\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONHR, -ARNH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1568\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1622\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e=CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003en3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1442\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e\u0026nbsp;CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1494\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eN-H, O-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONHR, -ARNH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1566\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1796\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eO-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-WATER\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003en6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1576\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eN-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-ONH\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\"\u003e\n \u003cp\u003e1770\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\"\u003e\n \u003cp\u003eC-H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CH2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.651877133105803%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.67235494880546%\" valign=\"bottom\"\u003e\n \u003cp\u003e1954\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"19.965870307167236%\" valign=\"bottom\"\u003e\n \u003cp\u003eC=O\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"44.7098976109215%\" valign=\"bottom\"\u003e\n \u003cp\u003e-CO2R\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eIt can be observed that saturated fatty acids were strongly correlated with the 1452 and 1540 nm wavelengths which correspond to O-H bonds in the first overtone region, and also with the 1460-1520 nm wavelengths which correspond to N-H bonds in the first overtone region. Monounsaturated fatty acids showed the most significant \u0026beta; values at 1410 nm, a wavelength related to O-H bonds, and at wavelengths of 1480 to 1580 nm related to N-H bonds in the first overtone region. Other areas where coefficients with high values were observed were the wavelengths between 1166-1365 nm and 1600-1800 nm related to C-H bonds. High value coefficients appeared at wavelengths of above 1900 nm correlated with the C=O bond in the second overtone region. It should be noted that in the case of the MUFA summation calibration the last three bonds (C-H, C-O, and C=O) have the highest weight. In the case of PUFA, the \u0026beta; coefficients with the highest values were at wavelengths 1166, 1442, 1620, and 1770 associated with the C-H bonds. Other coefficients appeared at wavelengths 1440-1460, 1614, 1800, and 1938 nm in association with O-H bonds and at wavelengths of between 1500 and 1580 nm associated with N-H bonds in the first overtone region. In addition, it should be added that at 1954 nm there was a high \u0026beta; coefficient value for the omega 6 calibration related to the C=O bond located in the second overtone region. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAlthough several studies have addressed the prediction of the lipid profile and the sums of the different groups of total fatty acids according to the amount of unsaturation through NIR spectroscopy in fresh beef, pork and rabbit meat\u0026nbsp;(Gonz\u0026aacute;lez-Martı́n et al., 2005; Pla, Hern\u0026aacute;ndez, Ari\u0026ntilde;o, Ram\u0026iacute;rez \u0026amp; D\u0026iacute;az, 2007; Realini, Duckett \u0026amp; Windham, 2004)\u0026nbsp;and in subcutaneous fat from Iberian pork\u0026nbsp;(Gonz\u0026aacute;lez-Mart\u0026iacute;n et al., 2021; P\u0026eacute;rez-Mar\u0026iacute;n, De Pedro Sanz, Guerrero-Ginel \u0026amp; Garrido-Varo, 2009)\u0026nbsp;and obtained acceptable equations, previous research on the prediction of the lipid profile in cured products deriving from Iberian pork is scarce\u0026nbsp;(Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., 2011; Tejerina et al., 2018). The results of Tejerina et al.\u0026nbsp;(2018)\u0026nbsp;revealed that the salt content and palmitic and oleic fatty acid percentages could be predicted with RSQ values of \u0026gt; 0.8 in packaged Iberian ham. Furthermore, Fern\u0026aacute;ndez-Cabanas et al.\u0026nbsp;(2011)\u0026nbsp;evaluated the feasibility of NIRS technology to determine the fatty acid profile in cured sausages from the Iberian pig (\u003cem\u003esalchich\u0026oacute;n\u003c/em\u003e and \u003cem\u003echorizo\u003c/em\u003e); they obtained RSQs of between 0.77 and 0.94 for the major fatty acids (C16:0, C18:0, C18:1, C18:2, and C18:3) and for the sums by degree of unsaturation.\u003c/p\u003e\n\u003cp\u003eAs for the minority fatty acids, although it was possible to calibrate some of them such as C12:0, C14:0, C17:0, C17:0, C17:1, and C20:1 with RSQ values of between 0.62 and 0.82 in fresh intact Iberian pork loin\u0026nbsp;(Gonz\u0026aacute;lez-Martı́n et al., 2005), other studies have however reported determination coefficients of nearly zero for the calibrations developed in dry-cured sausages for minor constituents such as C12:0, C17:0, C17:0, C17:1; C20:0, and C20:1\u0026nbsp;(Fern\u0026aacute;ndez-Caban\u0026aacute;s et al., 2011)\u0026nbsp;or coefficients lower than 0.5 in minced and homogenized rabbit meat\u0026nbsp;(Pla et al., 2007). In general, the recording of intact samples presents limitations as the models are less accurate in homogenized and minced samples\u0026nbsp;(Cozzolino \u0026amp; Murray, 2002)\u0026nbsp;owing to the different organization of muscle fibers, physical and chemical characteristics, the meat cut, the intramuscular fat content, and the moisture among other factors\u0026nbsp;(Cozzolino, Murray, Scaife \u0026amp; Paterson 2000; Cozzolino \u0026amp; Murray, 2002).\u003c/p\u003e\n\u003cp\u003eThe correlation of the values obtained in the laboratory with regard to those predicted by NIR with the fiber optic probe for the calibration set has been plotted in Figure 2 for stearic, \u0026alpha;-linolenic, oleic, linoleic, AGP, and n3 fatty acids. It can be seen that the validation RSQ values were higher than 0.6 and that the SEP and SEP (C) values were similar. The graphs allow the observation of the linear relationship between the reference values and those predicted in the calibration set. As far as we know, there are no previous studies on the prediction of fatty acids in hams or shoulders with which to compare RSQ values.\u003c/p\u003e\n\u003cp\u003eIn order to check the robustness of the model, the equations obtained were applied to 30 samples other than those used for calibration for which the spectral record and analytical data of the lipid profile are available. This allowed comparison of the predicted values with reference data obtained by laboratory analysis. The prediction results for the validation set of samples confirmed that there were no significant differences between the predicted data and the laboratory value (p\u0026gt;0.05) for all fatty acids except for myristic acid (C14:0). The root mean standard error (RMSE) values were between 0.13 for (C18:3 n6) and 2.02 for the summation of MUFA.\u003c/p\u003e\n\u003cp\u003e3.3.2 Calibration equations by the ANN method\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs previously described, the first step for developing calibration equations using artificial neural networks was to test what was the most suitable algorithm by using the cost function reduction criterion. The algorithm optimization process determined that the Scaled Conjugate Gradient, the Gradient Descent, the Gradient Descent with Adaptive Learning Rate, and the Gradient Descent with Momentum did not provide good predictive results. Better results were obtained with the Bayesian Regularization and Quasi-Newton algorithms albeit at a very high computational time cost. Finally, the Levenberg-Marquardt (LM) algorithm was used as it allowed a greater number of networks with RSQ values above 0.7 and at a reasonable time cost. This same algorithm obtained the best results in the prediction of sensory parameters for raw cured products\u0026nbsp;(Hern\u0026aacute;ndez-Jim\u0026eacute;nez et al., 2020; Hern\u0026aacute;ndez-Ramos et al., 2020).\u003c/p\u003e\n\u003cp\u003eThe ANN was then optimized to determine the number of neurons in the hidden layer, which had to be between the size of the input and output layer and was determined empirically (Berry \u0026amp; Linoff, 1997; Boger \u0026amp; Guterman, 1997). For each parameter, architectures with a variable number of neurons in the hidden layer (between 1 and 25) were analyzed and each of the architectures was trained 300 times from initial weights randomly assigned from a known seed value. The optimal number of neurons in the hidden layer depended on the parameter or constituent, which means that a higher number of neurons was necessary for the prediction of saturated fatty acids while the rest of the constituents presented a variable number of neurons, highlighting the value of less than 5 neurons in oleic, erucic, linoleic, and PUFA fatty acids (Table 4).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 4. Number of neurons in the hidden layer, RSQ, Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) values of the best ANN architecture for ecah fatty acid and summatories\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"444\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003eNumber of neurons in the hidden layer\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\"\u003e\n \u003cp\u003eNetworks \u0026ge;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\"\u003e\n \u003cp\u003eNetworks \u0026ge;0.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\"\u003e\n \u003cp\u003eRSQ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\"\u003e\n \u003cp\u003eMSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\"\u003e\n \u003cp\u003eRMSE\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC14:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC16:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC18:0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.86\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eSFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003e\u0026sum; C16:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC18:1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e234\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e2.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.44\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC18:1 n7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC22:1 n9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e267\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e1721\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.07\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eMUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC18:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e267\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC18:3 n3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e1080\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.10\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003eC20:2 n6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.07\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003ePUFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e177\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.75\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003en3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e208\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e1507\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"17.117117117117118%\" valign=\"bottom\"\u003e\n \u003cp\u003en6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.8018018018018%\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.09009009009009%\" valign=\"bottom\"\u003e\n \u003cp\u003e162\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"7.882882882882883%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.108108108108109%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.90990990990991%\" valign=\"bottom\"\u003e\n \u003cp\u003e0.75\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eRSQ: Coefficient of Determination, MSE: Mean Square Error, RMSE: Root Mean Square Error\u003c/p\u003e\n\u003cp\u003eFurthermore, to perform this network optimization the total set of samples was divided into three sets; i.e. a training set with 104 samples, the validation set with 22 samples, and the test set with 22 samples. For each parameter individual training was carried out to obtain the predictive neural network as detailed above. During the training of the network, it has to be checked that overtraining (or overfitting) of the network is not taking place; in order to do so the 22 samples of the validation set which had not participated in the training were used. Finally, the performance of the network was evaluated with the remaining 22 samples not included in the training and validation sets so as to obtain the RSQ and RMSE values and thus allow the selection of the best network. The results show that in all cases RSQ values of \u0026gt; 0.5 were obtained with the figures ranging between 0.56 and 0.81. The correlation coefficients in the case of lipid profile prediction in rapeseed oil were slightly higher with RSQ values of \u0026gt; 0.7\u0026nbsp;(Rajković et al., 2023). For all summations (SFA, MUFA, PUFA, n6, and n3) the RSQ values were higher than 0.6 (Table 4). It should be noted that the fatty acids and summations which presented the highest RSQ coefficients, with values of between 0.7 and 0.81, were the fatty acids C18:1, C18:2 n6, C18:3 n3, C22:1 n9, and n3, which showed a higher correlation with the free-range feeding of Iberian pigs\u0026nbsp;(Hern\u0026aacute;ndez-Jim\u0026eacute;nez et al., 2021).\u003c/p\u003e\n\u003cp\u003eA parameter which can be taken into account to ascertain the capacity of the ANN to predict the components analyzed is the number of networks which give an RSQ value higher than that established as a reference\u0026nbsp;(Hern\u0026aacute;ndez-Ramos et al., 2020). In this case it was only possible to find networks with an RSQ value of \u0026gt; 0.65 for oleic, erucic, \u0026beta; and \u0026alpha;-linolenic, PUFA, n3 and n6 acids, and for some of them the number of networks is less than 10 which gives an idea of the difficulty of calibrating these components. Although it has previously been reported that the higher the number of neurons in the hidden layer and the number of training sessions the better the predictive capacity of the network\u0026nbsp;(Hern\u0026aacute;ndez-Ramos et al., 2020), the results obtained in this study showed that there was no direct relationship between the number of neurons in the hidden layer and the RSQ. The RMSEs showed values of between 0.065 and 1.444 with errors of less than 1 in the prediction of all polyunsaturated fatty acids and their respective summations. Lower RMSE values were observed for the prediction of C18:1, C22:1, and C18:3 n3 and higher values for C16:0, C18:0, and C18:2 fatty acids than those reported in other studies predicting the lipid profile by ANN\u0026nbsp;(Rajković et al., 2023).\u003c/p\u003e\n\u003cp\u003eFinally, Figure 3 shows the prediction plots and RSQ values for some of the calculated fatty acids and sum values. They show a satisfactory correlation between the reference values and the values predicted by the network.\u003c/p\u003e\n\u003cp\u003e3.4 Comparison of the MPLS and ANN results\u003c/p\u003e\n\u003cp\u003eIf the results obtained by the MPLS and ANN methods are compared it can be seen that by using neural networks, it was possible to predict the 15 constituents by applying the RSQ \u0026gt; 0.5 criterion; by MPLS it was only possible to predict 8 constituents under this criterion. Furthermore, the use of ANNs allowed the obtaining of 4 equations with an RSQ value of \u0026gt; 0.7 while with the application of the MPLS methodology it was possible to predict only 3 parameters with this criterion (Tables 2 and 4). The performance of ANN was significantly better compared with the multiple regression model (Fig 2 and Fig 3), which is probably due to the fact that it is a non-linear method. This result is in line with previous studies conducted for the prediction of sensory parameters of meat products\u0026nbsp;(Hern\u0026aacute;ndez-Jim\u0026eacute;nez et al., 2020), fruit firmness on harvesting\u0026nbsp;(Torkashvan, Ahmadi \u0026amp; Nikravesh, 2017), and cheese texture during the ripening process\u0026nbsp;(V\u0026aacute;squez et al., 2018). However, other authors have reported similar results in the predictive ability of the physicochemical parameters of fresh pork using ANN and MPLS\u0026nbsp;(Prevolnik, Čandek-Potokar, Novič.\u0026amp; \u0026Scaron;korjanc, 2009).\u003c/p\u003e\n\u003cp\u003eThe wavelengths which have had the greatest weight in the generation of the models by the MPLS and ANN regression methods were compared. The interpretation is simpler in the case of the MPLS regression method since according to the general linear model NIRS data (Y) can be modelled as\u0026nbsp;Y =\u0026nbsp;b\u003csub\u003e0\u003c/sub\u003e + b\u003csub\u003e1\u003c/sub\u003eC\u003csub\u003el1\u003c/sub\u003e + b\u003csub\u003e2\u003c/sub\u003eC\u003csub\u003el2\u003c/sub\u003e + b\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003el3\u003c/sub\u003e +\u0026hellip;\u0026hellip;+ b\u003csub\u003en\u003c/sub\u003eC\u003csub\u003eln\u003c/sub\u003e,\u0026nbsp;in which \u0026beta;\u0026nbsp;refers to the coefficients measuring the contribution of each wavelength to explaining Y and\u0026nbsp;C\u003csub\u003el1\u003c/sub\u003e,C\u003csub\u003el2\u003c/sub\u003e,C\u003csub\u003el3\u003c/sub\u003e,....C\u003csub\u003eln\u003c/sub\u003e, are the wavelengths. A higher \u0026beta; value therefore means a higher contribution of that wavelength to the final equation. In the case of ANN, the methodology previously described by Hern\u0026aacute;ndez-Ramos et al. (2020) was applied. The effect of each input wavelength on the modelled variables was assessed by analyzing the weight of each of the inputs for each of the neurons in the hidden layer, as this value could reflect the importance of that input in the correct classification or predictive ability of that neuron. Once all the weights (positive or negative) were calculated for each of the neurons in the hidden layer (the number of which depended on the predicted variable), the quadratic weight of each of the inputs was calculated, i.e. the sum of the squared weight of each of the n neurons in the hidden layer for each of the 451 wavelengths. This process was carried out for all fatty acids and the data obtained for some of the most important ones (C16:0, C18:1, and C18:2 n3) are shown in Figure 4. To compare the MPLS and ANN results in a simple way, the square of each of the \u0026beta; values obtained for each of the 451 wavelengths was also calculated and plotted for these three fatty acids.\u003c/p\u003e\n\u003cp\u003eFirst of all, for the MPLS calibration it can clearly be seen that it is not a single wavelength as shown in Table 3 but a band around those \u0026beta; which make a significant contribution to the calibration. In the case of the C16:0 fatty acid, the following bands can be observed for the MPLS method with major contributions located around 1170-1200, 1260-1280, 1460-1480, 1510, 1580, 1610, 1660, 1780-1800, and 1910-1930. As mentioned above, these would correspond to absorption bands in the first and second overtones associated with the C-H, N-H, O-H and C=O chemical bonds. For this fatty acid it was observed that there was a coincidence with the inputs with more weight for ANN in the wavelengths close to 1180, 1260, and 1650, which correspond to the absorption bands of C-H bonds, and the 1510 band which corresponds to the absorption of N-H bonds. In addition to these wavelengths, when ANN calibration was carried out high quadratic weights were observed for the 1100 to 1300 bands, which are related to C-H and N-H bonds, for the 1380 to 1440 range in which the main absorbing bonds are C-H and O-H, for the wavelengths between 1540 and 1590 which are associated with O-H bonds, and for the 1660-1760 range corresponding to the first overtone of the C-H bond. The C-H, N-H and O-H bonds were therefore the most relevant for the calibration of this fatty acid as was previously observed in the case of MPLS.\u003c/p\u003e\n\u003cp\u003eAs far as C18:1 fatty acid is concerned, in addition to the bands around the \u0026beta; values previously indicated for the calibration with MPLS at 1410, 1722, 1746, and 1918 nm (Table 3), which are also important in calibration with ANN, this graph allows us to observe that there were other relevant bands in the range of 1110 to 1120 nm and around 1200 nm which correspond to the absorbance of C-H and O-H bonds and in the 1260-1280 interval. In addition to these, the bands around 1380 to 1590, 1870 to 1900, and 1940 to 1960 are of great importance in ANN calibration. These bands correspond to the absorbance of C-H, O-H, and N-H bonds in the first interval, of O-H and C=O bonds in the second interval, and of O-H, C=O, and N-H bonds in the last of the above-mentioned regions. These results confirm what was previously observed in MPLS, i.e. the important contribution of C=O bonds in the calibration of unsaturated fatty acids.\u003c/p\u003e\n\u003cp\u003eFinally, Figure 4c presents the results for the C18:3 n6 to show a smaller number of \u0026beta;-values which are relevant in MPLS calibration and narrower bands in both methods. In the MPLS method it can be observed that the bands with the highest contribution to the model development are located at 1170, 1440, 1500 to 1510, and 1780 as already discussed, and also that there is a significant contribution from the bands at 1120, 1210, and again in the 1260-1280 range, all of which are associated with the first and second overtone of the C-H bond. These bands do not correspond to those of greater weight in the ANN method, in which the most important contribution was distributed throughout the entire spectral range between 1380 and 2000 nm with narrow bands. The contributions found at 1610, 1720, 1820, and 1910 nm stand out, which as indicated in Table 3 correspond to the first overtone of C-H, to the C-O bond in oils, to the O-H bond, and to the C=O bond respectively.\u003c/p\u003e\n\u003cp\u003eIt should therefore be noted that in both methods the wavelengths with the highest weights are associated with the C-H, N-H, and O-H bonds, together with the C=O bond for unsaturated fatty acids. However, the coefficients in the MPLS method follow a distribution in defined bands over the whole spectral range, while in the case of ANN their behavior was very different depending on the component to be calibrated. Thus in some of the cases the highest weights appeared in the form of defined bands such as for oleic and linoleic acids. It should be noted that these fatty acids had a small number of neurons in the hidden layer (4), which probably explains this discrete distribution. However, for others such as palmitic acid weights are generally higher and more homogeneously distributed throughout the spectral range, which could be associated with the higher number of neurons in the hidden layer. Finally, the different assignment of weights to the different wavelengths, based on the different algorithms used, is probably the reason for the higher predictive ability of ANNs.\u003c/p\u003e"},{"header":"4. Conclusion","content":"\u003cp\u003eThe results obtained in this study indicate that it is possible to predict the main fatty acids and their summations calculated according to the degree of unsaturation in samples of Iberian hams and shoulders using non-destructive and rapid techniques such as NIR spectroscopy, which would facilitate the nutritional evaluation of the products and their labelling. Of the two chemometric methods used, artificial neural networks were able to predict a greater number of parameters with higher RSQ values than the MPLS method. These results indicate that in the prediction of the percentage of fatty acids in a complex matrix such as cured Iberian pork non-linear methods such as neural networks provided better results. However, this research could be continued by extending it with a larger number of samples in order to generate more robust models and confirm the behavior observed for both predictive methods.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interest\u003c/h2\u003e \u003cp\u003eFinancial interests: The authors declare they have no financial interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding:\u003c/h2\u003e \u003cp\u003eThis research was funded by Salamanca County Council (Spain) under grant number 18VEUH 463AC06.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eMiriam Hern\u0026aacute;ndez-Jim\u0026eacute;nez and Pedro Hern\u0026aacute;ndez Ramos: formal analysis, investigation, visualization. Isabel Revilla conceptualization, supervision, project administration, funding acquisition. Ana Vivar-Quintana: funding acquisition, conceptualization, editing. All the authors wrote and reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe are grateful to \u0026ldquo;Carrasco Ib\u0026eacute;ricos\u0026rdquo; Guijuelo (Salamanca) for their collaboration and Hern\u0026aacute;ndez-Jim\u0026eacute;nez M. is grateful to the Own Program III: Grants for Pre-doctoral Contracts of the University of Salamanca co-funded by Banco Santander.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlhamdan, A. M., Fickak, A., \u0026amp; Atia, A. R. (2019). Evaluation of sensory and texture profile analysis properties of stored Khalal Barhi dates nondestructively using Vis/NIR spectroscopy. \u003cem\u003eJournal of Food Process Engineering\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(6), e13215. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jfpe.13215\u003c/span\u003e\u003cspan address=\"10.1111/jfpe.13215\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlomar, D., Gallo, C., Casta\u0026ntilde;eda, M., \u0026amp; Fuchslocher, R. (2003). Chemical and discriminant analysis of bovine meat by near infrared reflectance spectroscopy (NIRS). \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e(4), 441\u0026ndash;450. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(02)00101-8\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(02)00101-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndr\u0026eacute;s, A. I., Cava, R., Mayoral, A. I., Tejeda, J. F., Morcuende, D., \u0026amp; Ruiz, J. (2001). Oxidative stability and fatty acid composition of pig muscles as affected by rearing system, crossbreeding and metabolic type of muscle fibre. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e59\u003c/em\u003e(1), 39\u0026ndash;47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(01)00050-X\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(01)00050-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndr\u0026eacute;s, S., Murray, I., Navajas, E. A., Fisher, A. V., Lambe, N. R., \u0026amp; B\u0026uuml;nger, L. (2007). Prediction of sensory characteristics of lamb meat samples by near infrared reflectance spectroscopy. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e76\u003c/em\u003e(3), 509\u0026ndash;516. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2007.01.011\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2007.01.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnsorena, D., \u0026amp; Astiasar\u0026aacute;n, I. (2004). Effect of storage and packaging on fatty acid composition and oxidation in dry fermented sausages made with added olive oil and antioxidants. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e67\u003c/em\u003e(2), 237\u0026ndash;244. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2003.10.011\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2003.10.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAntequera, T., L\u0026oacute;pez-Bote, C. J., C\u0026oacute;rdoba, J. J., Garc\u0026iacute;a, C., Asensio, M. A., Ventanas, J., Garc\u0026iacute;a-Regueiro, J. A., \u0026amp; D\u0026iacute;az, I. (1992). Lipid oxidative changes in the processing of Iberian pig hams. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(2), 105\u0026ndash;110. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0308-8146(92)90018-W\u003c/span\u003e\u003cspan address=\"10.1016/0308-8146(92)90018-W\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArce, L., Dom\u0026iacute;nguez-Vidal, A., Rodr\u0026iacute;guez-Est\u0026eacute;vez, V., L\u0026oacute;pez-Vidal, S., Ayora-Ca\u0026ntilde;ada, M. J., \u0026amp; Valc\u0026aacute;rcel, M. (2009). Feasibility study on the use of infrared spectroscopy for the direct authentication of Iberian pig fattening diet. \u003cem\u003eAnalytica Chimica Acta\u003c/em\u003e, \u003cem\u003e636\u003c/em\u003e(2), 183\u0026ndash;189. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aca.2009.01.058\u003c/span\u003e\u003cspan address=\"10.1016/j.aca.2009.01.058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAristoy, M. C., \u0026amp; Toldr\u0026aacute;, F. (1998). Concentration of free amino acids and dipeptides in porcine skeletal muscles with different oxidative patterns. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(3), 327\u0026ndash;332. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(98)00037-0\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(98)00037-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalabin, R. M., \u0026amp; Safieva, R. Z. (2011). Near-Infrared (NIR) Spectroscopy for Biodiesel Analysis: Fractional Composition, Iodine Value, and Cold Filter Plugging Point from One Vibrational Spectrum. \u003cem\u003eEnergy \u0026amp; Fuels\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(5), 2373\u0026ndash;2382. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1021/ef200356h\u003c/span\u003e\u003cspan address=\"10.1021/ef200356h\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalabin, R. M., Safieva, R. Z., \u0026amp; Lomakina, E. I. (2007). Comparison of linear and nonlinear calibration models based on near infrared (NIR) spectroscopy data for gasoline properties prediction. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e, \u003cem\u003e88\u003c/em\u003e(2), 183\u0026ndash;188. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chemolab.2007.04.006\u003c/span\u003e\u003cspan address=\"10.1016/j.chemolab.2007.04.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalabin, R. M., Safieva, R. Z., \u0026amp; Lomakina, E. I. (2008). Wavelet neural network (WNN) approach for calibration model building based on gasoline near infrared (NIR) spectra. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e, \u003cem\u003e93\u003c/em\u003e(1), 58\u0026ndash;62. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chemolab.2008.04.003\u003c/span\u003e\u003cspan address=\"10.1016/j.chemolab.2008.04.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarlocco, N., Vadell, A., Ballesteros, F., Galietta, G., \u0026amp; Cozzolino, D. (2006). Predicting intramuscular fat, moisture and Warner-Bratzler shear force in pork muscle using near infrared reflectance spectroscopy. \u003cem\u003eAnimal Science\u003c/em\u003e, \u003cem\u003e82\u003c/em\u003e(1), 111\u0026ndash;116. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1079/ASC20055\u003c/span\u003e\u003cspan address=\"10.1079/ASC20055\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerry, M. J. A., \u0026amp; Linoff, G. S. (1997). \u003cem\u003eData Mining Techniques: For Marketing, Sales, and Customer Relationship Management\u003c/em\u003e. Wiley.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoger, Z., \u0026amp; Guterman, H. (1997). Knowledge extraction from artificial neural network models. \u003cem\u003eComputational Cybernetics and Simulation 1997 IEEE International Conference on Systems, Man, and Cybernetics\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e, 3030\u0026ndash;3035 vol.4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/ICSMC.1997.633051\u003c/span\u003e\u003cspan address=\"10.1109/ICSMC.1997.633051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuscailhon, S., \u0026amp; Monin, G. (1994). D\u0026eacute;terminisme des qualit\u0026eacute;s sensorielles du jambon sec. \u003cem\u003eD\u0026eacute;terminisme Des Qualit\u0026eacute;s Sensorielles Du Jambon Sec\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(1), 23\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarrapiso, A. I., Bonilla, F., \u0026amp; Garcı́a, C. (2003). Effect of crossbreeding and rearing system on sensory characteristics of Iberian ham. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e65\u003c/em\u003e(1), 623\u0026ndash;629. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(02)00256-5\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(02)00256-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCava, R., Est\u0026eacute;vez, M., Morcuende, D., \u0026amp; Antequera, T. (2003). Evolution of fatty acids from intramuscular lipid fractions during ripening of Iberian hams as affected by α-tocopheryl acetate supplementation in diet. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e(2), 199\u0026ndash;207. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0308-8146(02)00413-2\u003c/span\u003e\u003cspan address=\"10.1016/S0308-8146(02)00413-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCava, R., Ruiz, J., L\u0026oacute;pez-Bote, C., Mart\u0026iacute;n, L., Garc\u0026iacute;a, C., Ventanas, J., \u0026amp; Antequera, T. (1997). Influence of finishing diet on fatty acid profiles of intramuscular lipids, triglycerides and phospholipids in muscles of the Iberian pig. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(2), 263\u0026ndash;270. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(96)00102-7\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(96)00102-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCozzolino, D., \u0026amp; Murray, I. (2002). Effect of Sample Presentation and Animal Muscle Species on the Analysis of Meat by near Infrared Reflectance Spectroscopy. \u003cem\u003eJournal of Near Infrared Spectroscopy\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(1), 37\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCozzolino, D., Murray, I., Scaife, J. R., \u0026amp; Paterson, R. (2000). Study of dissected lamb muscles by visible and near infrared reflectance spectroscopy for composition assessment. \u003cem\u003eAnimal Science\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(3), 417\u0026ndash;423.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Maesschalck, R., Jouan-Rimbaud, D., \u0026amp; Massart, D. L. (2000). The Mahalanobis distance. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(1), 1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0169-7439(99)00047-7\u003c/span\u003e\u003cspan address=\"10.1016/S0169-7439(99)00047-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFern\u0026aacute;ndez-Caban\u0026aacute;s, V. M., Garrido-Varo, A., Garc\u0026iacute;a Olmo, J., De Pedro, E., \u0026amp; Dardenne, P. (2007). Optimisation of the spectral pre-treatments used for Iberian pig fat NIR calibrations. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e, \u003cem\u003e87\u003c/em\u003e(1), 104\u0026ndash;112. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chemolab.2006.10.005\u003c/span\u003e\u003cspan address=\"10.1016/j.chemolab.2006.10.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFern\u0026aacute;ndez-Caban\u0026aacute;s, V. M., Polvillo, O., Rodr\u0026iacute;guez-Acu\u0026ntilde;a, R., Botella, B., \u0026amp; Horcada, A. (2011). Rapid determination of the fatty acid profile in pork dry-cured sausages by NIR spectroscopy. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e124\u003c/em\u003e(1), 373\u0026ndash;378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2010.06.031\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2010.06.031\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFolch, J., Lees, M., \u0026amp; Sloane Stanley, G. H. (1957). A simple method for the isolation and purification of total lipids from animal tissues. \u003cem\u003eJournal Of Biological Chemistry\u003c/em\u003e, \u003cem\u003e226\u003c/em\u003e(1), 497\u0026ndash;509.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGandemer, G. (2002). Lipids in muscles and adipose tissues, changes during processing and sensory properties of meat products. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e62\u003c/em\u003e(3), 309\u0026ndash;321. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(02)00128-6\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(02)00128-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarc\u0026iacute;a, A. I. M., Andrada, A. D., Mu\u0026ntilde;oz, A. I. R., \u0026amp; Bote, C. J. L. (2005). Effect of feeding level during the period previous to free range fattening on growth and carcass characteristics. \u003cem\u003eSpanish Journal of Agricultural Research\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e, 387\u0026ndash;395.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarc\u0026iacute;a-Olmo, J., Garrido-Varo, A., \u0026amp; De Pedro, E. (2001). The Transfer of Fatty Acid Calibration Equations Using Four Sets of Unsealed Liquid Standardisation Samples. \u003cem\u003eJournal of Near Infrared Spectroscopy\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(1), 49\u0026ndash;62. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1255/jnirs.294\u003c/span\u003e\u003cspan address=\"10.1255/jnirs.294\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarrido-Varo, A., Carrete, R., \u0026amp; Fern\u0026aacute;ndez-Caban\u0026aacute;s, V. (1998). Use of Difference near Infrared Reflectance Spectra to Extract Relevant Information from the Spectra of Agro\u0026ndash;Food Products. \u003cem\u003eJournal of Near Infrared Spectroscopy\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(1), 89\u0026ndash;95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1255/jnirs.125\u003c/span\u003e\u003cspan address=\"10.1255/jnirs.125\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez, E., \u0026amp; Tejeda, J. F. (2007). Effects of dietary incorporation of different antioxidant extracts and free-range rearing on fatty acid composition and lipid oxidation of Iberian pig meat. \u003cem\u003eAnimal\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(7), 1060\u0026ndash;1067. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/S1751731107000195\u003c/span\u003e\u003cspan address=\"10.1017/S1751731107000195\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Mart\u0026iacute;n, M. I., Escuredo, O., Hern\u0026aacute;ndez-Jim\u0026eacute;nez, M., Revilla, I., Vivar-Quintana, A. M., Mart\u0026iacute;nez-Mart\u0026iacute;n, I., \u0026amp; Hern\u0026aacute;ndez-Ramos, P. (2021). Prediction of stable isotopes and fatty acids in subcutaneous fat of Iberian pigs by means of NIR: A comparison between benchtop and portable systems. \u003cem\u003eTalanta\u003c/em\u003e, \u003cem\u003e224\u003c/em\u003e, 121817. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.talanta.2020.121817\u003c/span\u003e\u003cspan address=\"10.1016/j.talanta.2020.121817\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Mart\u0026iacute;n, M. I., Palacios, V., Revilla, V., Vivar-Quintana, I., A. M., \u0026amp; Hern\u0026aacute;ndez-Hierro, J. M. (2017). Discrimination between cheeses made from cow\u0026rsquo;s, ewe\u0026rsquo;s and goat\u0026rsquo;s milk from unsaturated fatty acids and use of the canonical biplot method. \u003cem\u003eJournal of Food Composition and Analysis\u003c/em\u003e, \u003cem\u003e56\u003c/em\u003e, 34\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfca.2016.12.005\u003c/span\u003e\u003cspan address=\"10.1016/j.jfca.2016.12.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Martı́n, I., Gonz\u0026aacute;lez-P\u0026eacute;rez, C., Alvarez-Garcı́a, N., \u0026amp; Gonz\u0026aacute;lez-Cabrera, J. M. (2005). On-line determination of fatty acid composition in intramuscular fat of Iberian pork loin by NIRs with a remote reflectance fibre optic probe. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e69\u003c/em\u003e(2), 243\u0026ndash;248. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2004.07.003\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2004.07.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenckel, P., Oksbjerg, N., Erlandsen, E., Barton-Gade, P., \u0026amp; Bejerholm, C. (1997). Histo- and biochemical characteristics of the Longissimus dorsi muscle in pigs and their relationships to performance and meat quality. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(3), 311\u0026ndash;321. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(97)00063-6\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(97)00063-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHern\u0026aacute;ndez-Jim\u0026eacute;nez, M., Gonz\u0026aacute;lez-Mart\u0026iacute;n, M. I., Mart\u0026iacute;nez-Mart\u0026iacute;n, I., Revilla, I., \u0026amp; Vivar-Quintana, A. M. (2021). Carbon stable isotopes, fatty acids and the use of NIRS to differentiate IBERIAN pigs. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e182\u003c/em\u003e, 108619. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2021.108619\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2021.108619\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHern\u0026aacute;ndez-Jim\u0026eacute;nez, M., Hern\u0026aacute;ndez-Ramos, P., Mart\u0026iacute;nez-Mart\u0026iacute;n, I., Vivar-Quintana, A. M., Gonz\u0026aacute;lez-Mart\u0026iacute;n, I., \u0026amp; Revilla, I. (2020). Comparison of artificial neural networks and multiple regression tools applied to near infrared spectroscopy for predicting sensory properties of products from quality labels. \u003cem\u003eMicrochemical Journal\u003c/em\u003e, \u003cem\u003e159\u003c/em\u003e, 105459. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.microc.2020.105459\u003c/span\u003e\u003cspan address=\"10.1016/j.microc.2020.105459\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHern\u0026aacute;ndez-Ramos, P., Vivar-Quintana, A. M., Revilla, I., Gonz\u0026aacute;lez-Mart\u0026iacute;n, M. I., Hern\u0026aacute;ndez-Jim\u0026eacute;nez, M., \u0026amp; Mart\u0026iacute;nez-Mart\u0026iacute;n, I. (2020). Prediction of Sensory Parameters of Cured Ham: A Study of the Viability of the Use of NIR Spectroscopy and Artificial Neural Networks. \u003cem\u003eSensors (Basel, Switzerland)\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(19). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s20195624\u003c/span\u003e\u003cspan address=\"10.3390/s20195624\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Article 19.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorcada, A., Valera, M., Ju\u0026aacute;rez, M., \u0026amp; Fern\u0026aacute;ndez-Caban\u0026aacute;s, V. M. (2020). Authentication of Iberian pork official quality categories using a portable near infrared spectroscopy (NIRS) instrument. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e318\u003c/em\u003e, 126\u0026ndash;133. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2020.126471\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2020.126471\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHruschaka, W. (2001). \u003cem\u003eData analysis: Wavelengt selection methods\u003c/em\u003e (P. C. Williams \u0026amp; N. Karl, Eds.). Near infrared technology in the agricultural and food industries, American Association of Cereal Chemist.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJurinjak Tušek, A., Jurina, T., Čulo, I., Valinger, D., Gajdoš Kljusurić, J., \u0026amp; Benković, M. (2022). Application of NIRs coupled with PLS and ANN modelling to predict average droplet size in oil-in-water emulsions prepared with different microfluidic devices. \u003cem\u003eSpectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy\u003c/em\u003e, \u003cem\u003e270\u003c/em\u003e, 120860. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.saa.2022.120860\u003c/span\u003e\u003cspan address=\"10.1016/j.saa.2022.120860\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKundu, P., Paul, V., Kumar, V., \u0026amp; Mishra, I. M. (2015). Formulation development, modeling and optimization of emulsification process using evolving RSM coupled hybrid ANN-GA framework. \u003cem\u003eChemical Engineering Research and Design\u003c/em\u003e, \u003cem\u003e104\u003c/em\u003e, 773\u0026ndash;790. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cherd.2015.10.025\u003c/span\u003e\u003cspan address=\"10.1016/j.cherd.2015.10.025\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeroy, B., Lambotte, S., Dotreppe, O., Lecocq, H., Istasse, L., \u0026amp; Clinquart, A. (2004). Prediction of technological and organoleptic properties of beef \u003cem\u003eLongissimus thoracis\u003c/em\u003e from near-infrared reflectance and transmission spectra. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e66\u003c/em\u003e(1), 45\u0026ndash;54. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(03)00002-0\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(03)00002-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLucas, A., Andueza, D., Ferlay, A., \u0026amp; Martin, B. (2008). Prediction of fatty acid composition of fresh and freeze-dried cheeses by visible\u0026ndash;near-infrared reflectance spectroscopy. \u003cem\u003eInternational Dairy Journal\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(6), 595\u0026ndash;604. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.idairyj.2007.12.001\u003c/span\u003e\u003cspan address=\"10.1016/j.idairyj.2007.12.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarini, F. (2013). In M. Bevilacqua, R. Bucci, A. D. Magr\u0026igrave;, A. L. Magr\u0026igrave;, R. Nescatelli, \u0026amp; F. Marini (Eds.), \u003cem\u003eChemometrics in food chemistry Chap. 5. Classification and class-modelling\u003c/em\u003e (Vol. 1,281st ed., 28).). Elsevier Inc.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMart\u0026iacute;nez-Mart\u0026iacute;n, I., Hern\u0026aacute;ndez-Jim\u0026eacute;nez, M., Revilla, I., \u0026amp; Vivar-Quintana, A. M. (2023). Prediction of Mineral Composition in Wheat Flours Fortified with Lentil Flour Using NIR Technology. \u003cem\u003eSensors (Basel, Switzerland)\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 3. https://doi.org/10.3390/s23031491\u003c/span\u003e\u003cspan address=\"Article 3. 10.3390/s23031491\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMart\u0026iacute;n-C\u0026aacute;ceres, L. (1996). \u003cem\u003eInfluencia de las condiciones del procesado sobre los cambios madurativos en el jam\u0026oacute;n ib\u0026eacute;rico\u003c/em\u003e [Tesis docotral, Universidad de Extremadura].\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMart\u0026iacute;n-G\u0026oacute;mez, A., Arroyo-Manzanares, N., Rodr\u0026iacute;guez-Est\u0026eacute;vez, V., \u0026amp; Arce, L. (2019). Use of a non-destructive sampling method for characterization of Iberian cured ham breed and feeding regime using GC-IMS. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e152\u003c/em\u003e, 146\u0026ndash;154. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2019.02.018\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2019.02.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Scopus.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMataix, J., Quiles, J. L., \u0026amp; Rodr\u0026iacute;guez, J. (2001). Aporte de grasa. \u003cem\u003eGu\u0026iacute;as Alimentarias Para La Poblaci\u0026oacute;n Espa\u0026ntilde;ola Recomendaciones Para Una Dieta Saludable Madrid: SENC\u003c/em\u003e, 231\u0026ndash;237.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNorris, K. H. (2001). Understanding and Correcting the Factors Which Affect Diffuse Transmittance Spectra. \u003cem\u003eNIR News\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(3), 6\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1255/nirn.613\u003c/span\u003e\u003cspan address=\"10.1255/nirn.613\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrtiz, A., Le\u0026oacute;n, L., Contador, R., \u0026amp; Tejerina, D. (2021). Near-Infrared Spectroscopy (NIRS) as a Tool for Classification of Pre-Sliced Iberian Salchich\u0026oacute;n, Modified Atmosphere Packaged (MAP) According to the Official Commercial Categories of Raw Meat. \u003cem\u003eFoods\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(8). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/foods10081865\u003c/span\u003e\u003cspan address=\"10.3390/foods10081865\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Article 8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePajuelo, A., S\u0026aacute;nchez, S., P\u0026eacute;rez-Palacios, T., Caballero, D., D\u0026iacute;az, J., Antequera, T., \u0026amp; Marcos, C. F. (2022). 1H NMR to analyse the lipid profile in the glyceride fraction of different categories of Iberian dry-cured hams. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e383\u003c/em\u003e, 132371. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2022.132371\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2022.132371\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026eacute;rez-Mar\u0026iacute;n, D., De Pedro Sanz, E., Guerrero-Ginel, J. E., \u0026amp; Garrido-Varo, A. (2009). A feasibility study on the use of near-infrared spectroscopy for prediction of the fatty acid profile in live Iberian pigs and carcasses. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e83\u003c/em\u003e(4), 627\u0026ndash;633. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2009.07.012\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2009.07.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026eacute;rez-Mar\u0026iacute;n, D., Fearn, T., Riccioli, C., De Pedro, E., \u0026amp; Garrido, A. (2021). Probabilistic classification models for the in situ authentication of iberian pig carcasses using near infrared spectroscopy. \u003cem\u003eTalanta\u003c/em\u003e, \u003cem\u003e222\u003c/em\u003e, 121511. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.talanta.2020.121511\u003c/span\u003e\u003cspan address=\"10.1016/j.talanta.2020.121511\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026eacute;rez-Mar\u0026iacute;n, D., Garrido-Varo, A., De Pedro, E., \u0026amp; Guerrero-Ginel, J. E. (2007). Chemometric utilities to achieve robustness in liquid NIRS calibrations: Application to pig fat analysis. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e, \u003cem\u003e87\u003c/em\u003e(2), 241\u0026ndash;246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chemolab.2007.02.004\u003c/span\u003e\u003cspan address=\"10.1016/j.chemolab.2007.02.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePillonel, L., Badertscher, R., Casey, M., Meyer, J., Rossmann, A., Schlichtherle-Cerny, H., Tabacchi, R., \u0026amp; Bosset, J. O. (2005). Geographic origin of European Emmental cheese: Characterisation and descriptive statistics. \u003cem\u003eInternational Dairy Journal\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(6), 547\u0026ndash;556. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.idairyj.2004.07.028\u003c/span\u003e\u003cspan address=\"10.1016/j.idairyj.2004.07.028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePla, M., Hern\u0026aacute;ndez, P., Ari\u0026ntilde;o, B., Ram\u0026iacute;rez, J. A., \u0026amp; D\u0026iacute;az, I. (2007). Prediction of fatty acid content in rabbit meat and discrimination between conventional and organic production systems by NIRS methodology. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e100\u003c/em\u003e(1), 165\u0026ndash;170. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2005.09.029\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2005.09.029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrevolnik, M., Čandek-Potokar, M., Novič, M., \u0026amp; Škorjanc, D. (2009). An attempt to predict pork drip loss from pH and colour measurements or near infrared spectra using artificial neural networks. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e83\u003c/em\u003e(3), 405\u0026ndash;411. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2009.06.015\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2009.06.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrieto, N., Andr\u0026eacute;s, S., Gir\u0026aacute;ldez, F. J., Mantec\u0026oacute;n, A. R., \u0026amp; Lav\u0026iacute;n, P. (2008). Ability of near infrared reflectance spectroscopy (NIRS) to estimate physical parameters of adult steers (oxen) and young cattle meat samples. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e79\u003c/em\u003e(4), 692\u0026ndash;699. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2007.10.035\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2007.10.035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrieto, N., Pawluczyk, O., Dugan, M. E. R., \u0026amp; Aalhus, J. L. (2017). A Review of the Principles and Applications of Near-Infrared Spectroscopy to Characterize Meat, Fat, and Meat Products. \u003cem\u003eApplied Spectroscopy\u003c/em\u003e, \u003cem\u003e71\u003c/em\u003e(7), 1403\u0026ndash;1426. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0003702817709299\u003c/span\u003e\u003cspan address=\"10.1177/0003702817709299\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRajković, D., Marjanović Jeromela, A., Pezo, L., Lončar, B., Grahovac, N., \u0026amp; Kondić Špika, A. (2023). Artificial neural network and random forest regression models for modelling fatty acid and tocopherol content in oil of winter rapeseed. \u003cem\u003eJournal of Food Composition and Analysis\u003c/em\u003e, \u003cem\u003e115\u003c/em\u003e, 105020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfca.2022.105020\u003c/span\u003e\u003cspan address=\"10.1016/j.jfca.2022.105020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRD 4/ (2014). de 10 de enero, por el que se aprueba la norma de calidad para la carne, el jam\u0026oacute;n, la paleta y la ca\u0026ntilde;a de lomo ib\u0026eacute;rico.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRealini, C. E., Duckett, S. K., \u0026amp; Windham, W. R. (2004). Effect of vitamin C addition to ground beef from grass-fed or grain-fed sources on color and lipid stability, and prediction of fatty acid composition by near-infrared reflectance analysis. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e68\u003c/em\u003e(1), 35\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2004.02.002\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2004.02.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRipoll, G., Albert\u0026iacute;, P., Panea, B., Olleta, J. L., \u0026amp; Sa\u0026ntilde;udo, C. (2008). Near-infrared reflectance spectroscopy for predicting chemical, instrumental and sensory quality of beef. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e80\u003c/em\u003e(3), 697\u0026ndash;702. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2008.03.009\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2008.03.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodr\u0026iacute;guez-Est\u0026eacute;vez, V., Garc\u0026iacute;a, A., Pe\u0026ntilde;a, F., \u0026amp; G\u0026oacute;mez, A. G. (2009). Foraging of Iberian fattening pigs grazing natural pasture in the dehesa. \u003cem\u003eLivestock Science\u003c/em\u003e, \u003cem\u003e120\u003c/em\u003e(1), 135\u0026ndash;143. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.livsci.2008.05.006\u003c/span\u003e\u003cspan address=\"10.1016/j.livsci.2008.05.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRubio, B., Mart\u0026iacute;nez, B., S\u0026aacute;nchez, M. J., Garc\u0026iacute;a-Cach\u0026aacute;n, M. D., Rovira, J., \u0026amp; Jaime, I. (2007). Study of the shelf life of a dry fermented sausage salchichon made from raw material enriched in monounsaturated and polyunsaturated fatty acids and stored under modified atmospheres. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e76\u003c/em\u003e(1), 128\u0026ndash;137. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2006.10.021\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2006.10.021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRuiz-Carrascal, J., Ventanas, J., Cava, R., Andr\u0026eacute;s, A. I., \u0026amp; Garcı́a, C. (2000). Texture and appearance of dry cured ham as affected by fat content and fatty acid composition. \u003cem\u003eFood Research International\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(2), 91\u0026ndash;95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0963-9969(99)00153-2\u003c/span\u003e\u003cspan address=\"10.1016/S0963-9969(99)00153-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShahidi, F., \u0026amp; Zhong, Y. (2010). Lipid oxidation and improving the oxidative stability. \u003cem\u003eChemical Society Reviews\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e(11), 4067\u0026ndash;4079. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1039/B922183M\u003c/span\u003e\u003cspan address=\"10.1039/B922183M\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheard, P. R., Enser, M., Wood, J. D., Nute, G. R., Gill, B. P., \u0026amp; Richardson, R. I. (2000). Shelf life and quality of pork and pork products with raised n-3 PUFA. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e(2), 213\u0026ndash;221. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(99)00145-X\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(99)00145-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShenk, J. S., \u0026amp; Westerhaus, M. O. (1996). Calibration the ISI way. \u003cem\u003eNear Infrared Spectroscopy: The Future Waves\u003c/em\u003e, 198\u0026ndash;202.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTejeda, J. F., Gandemer, G., Antequera, T., Viau, M., \u0026amp; Garcı́a, C. (2002). Lipid traits of muscles as related to genotype and fattening diet in Iberian pigs: Total intramuscular lipids and triacylglycerols. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e60\u003c/em\u003e(4), 357\u0026ndash;363. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0309-1740(01)00143-7\u003c/span\u003e\u003cspan address=\"10.1016/S0309-1740(01)00143-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTejerina, D., Garc\u0026iacute;a Torres, S., Cabeza de Vaca, M., Ortiz, A., \u0026amp; Romero, M. P. (2018). Evaluation of near-infrared spectroscopy (NIRS) for the quality control of packaged cured ham-sliced from Iberian pigs. \u003cem\u003eArchivos de Zootecnia\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e, 231\u0026ndash;234. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21071/az.v67iSupplement.3611\u003c/span\u003e\u003cspan address=\"10.21071/az.v67iSupplement.3611\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToldr\u0026aacute;, F., \u0026amp; Aristoy, M. C. (2004). Dry-cured ham. \u003cem\u003eFOOD SCIENCE AND TECHNOLOGY-NEW YORK-MARCEL DEKKER-\u003c/em\u003e, 369\u0026ndash;384.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToldr\u0026aacute;, F., \u0026amp; Flores, M. (1998). The Role of Muscle Proteases and Lipases in Flavor Development During the Processing of Dry-Cured Ham. \u003cem\u003eCritical Reviews in Food Science and Nutrition\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(4), 331\u0026ndash;352. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10408699891274237\u003c/span\u003e\u003cspan address=\"10.1080/10408699891274237\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTorkashvand, A. M., Ahmadi, A., \u0026amp; Nikravesh, N. L. (2017). Prediction of kiwifruit firmness using fruit mineral nutrient concentration by artificial neural network (ANN) and multiple linear regressions (MLR). \u003cem\u003eJournal of Integrative Agriculture\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(7), 1634\u0026ndash;1644. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S2095-3119(16)61546-0\u003c/span\u003e\u003cspan address=\"10.1016/S2095-3119(16)61546-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eV\u0026aacute;squez, N., Mag\u0026aacute;n, C., Oblitas, J., Chuquizuta, T., Avila-George, H., \u0026amp; Castro, W. (2018). Comparison between artificial neural network and partial least squares regression models for hardness modeling during the ripening process of Swiss-type cheese using spectral profiles. \u003cem\u003eJournal of Food Engineering\u003c/em\u003e, \u003cem\u003e219\u003c/em\u003e, 8\u0026ndash;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfoodeng.2017.09.008\u003c/span\u003e\u003cspan address=\"10.1016/j.jfoodeng.2017.09.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie, L. H., Tang, S. Q., Wei, X. J., Sheng, Z. H., Shao, G. N., Jiao, G. A., Hu, S. K., Wang-Lin, \u0026amp; Hu, P. S. (2022). Simultaneous determination of apparent amylose, amylose and amylopectin content and classification of waxy rice using near-infrared spectroscopy (NIRS). \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e388\u003c/em\u003e, 132944. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2022.132944\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2022.132944\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZamora-Rojas, E., Garrido-Varo, A., De Pedro-Sanz, E., Guerrero-Ginel, J. E., \u0026amp; P\u0026eacute;rez-Mar\u0026iacute;n, D. (2011). Monitoring NIRS calibrations for use in routine meat analysis as part of Iberian pig-breeding programs. \u003cem\u003eFood Chemistry\u003c/em\u003e, \u003cem\u003e129\u003c/em\u003e(4), 1889\u0026ndash;1897. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foodchem.2011.05.139\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2011.05.139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou, L. J., Wu, H., Li, J. T., Wang, Z. Y., \u0026amp; Zhang, L. Y. (2012). Determination of fatty acids in broiler breast meat by near-infrared reflectance spectroscopy. \u003cem\u003eMeat Science\u003c/em\u003e, \u003cem\u003e90\u003c/em\u003e(3), 658\u0026ndash;664. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.meatsci.2011.10.010\u003c/span\u003e\u003cspan address=\"10.1016/j.meatsci.2011.10.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"food-and-bioprocess-technology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Food and Bioprocess Technology](https://www.springer.com/journal/11947)","snPcode":"11947","submissionUrl":"https://submission.nature.com/new-submission/11947/3","title":"Food and Bioprocess Technology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4241621/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4241621/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn this study the feasibility of predicting the lipid profiles of Iberian ham and shoulder samples by using NIRS technology was evaluated. Gas chromatography analysis was the reference method used. The muscles analyzed and recorded by NIRS were 76 \u003cem\u003eBiceps femoris\u003c/em\u003e for Iberian hams and 72\u003cem\u003eBrachiocephalicus\u003c/em\u003e for Iberian shoulders. NIR calibrations were carried out by using two methods: modified partial least squares regression (MPLS) and artificial neural networks (ANN). With the MPLS method it was possible to obtain equations with RSQ of \u0026gt; 0.5 for 5 individual fatty acids and 3 summations (PUFA, n-3 and n-6). The use of neural networks made it possible to find equations with RSQ of \u0026gt; 0.5 for 10 individual fatty acids, all of which are present in over 90% of the samples, and 5 summates (SFA, MUFA, PUFA, n-3 and n-6); finding that the calibration curves of the fatty acids C18:1, C18:2n6 and C18:3n3 presented RSQs of \u0026gt; 0.7. The results obtained indicate that NIR spectroscopy could be a very useful technology for the quality control of cured products as it allows estimating the main fatty constituents quickly and without using reagents.\u003c/p\u003e","manuscriptTitle":"Prediction of the Fatty Acid Profiles of Iberian Pig Products by Near Infrared Spectroscopy: A Comparison Between Multiple Regression Tools and Artificial Neural Networks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-19 19:14:19","doi":"10.21203/rs.3.rs-4241621/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-05-29T09:45:41+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-27T13:07:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"68355852042806681823744108785673689134","date":"2024-05-23T11:12:05+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-07T09:14:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"189320586081330486472226024950397340781","date":"2024-05-03T12:36:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"b46beed5-19bd-4701-a89f-d182f2e069ec","date":"2024-04-19T02:18:41+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-04-17T02:15:54+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-04-15T06:57:27+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-04-15T00:39:59+00:00","index":"","fulltext":""},{"type":"submitted","content":"Food and Bioprocess Technology","date":"2024-04-09T11:45:37+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"food-and-bioprocess-technology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Food and Bioprocess Technology](https://www.springer.com/journal/11947)","snPcode":"11947","submissionUrl":"https://submission.nature.com/new-submission/11947/3","title":"Food and Bioprocess Technology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f6fd1ae5-f16b-46fa-b38c-13eed46e5479","owner":[],"postedDate":"April 19th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2024-06-15T04:54:22+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-19 19:14:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4241621","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4241621","identity":"rs-4241621","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}