Age Prediction based on Blood DNA Methylation

Objective In the judicial field, traditional DNA methylation age prediction models have low accuracy and poor stability. Additionally, the use of linear regression models for detection is inefficient and costly. This study aims to utilize the prediction principles of the Support Vector Regression (SVR) model, based on preliminary laboratory data from blood DNA methylation detection using the Illumina 850K chip. By selecting low-dimensional and highly linear loci, we aim to establish a highly stable and accurate blood DNA methylation age prediction model. Methods This research is based on Illumina 850K chip technology. We conducted a literature review to select CpG sites and related primers, then employed SVR for model construction and age prediction. The model was built on the Matlab2022a platform. Standard parameters were selected, and optimal values for C and g were determined using grid search and cross-validation methods. During data processing, numerical values were normalized before calculation and de-normalized to obtain the predicted values.

The constructed model achieved an R² of 0.91563 and a Mean Absolute Error (MAE) of 2.77 years. This indicates that the prediction accuracy for blood samples reached 91.56%, with an error of 2.77 years. Moreover, the accuracy of the model's predictions decreases with increasing age.

forensic genetics; SVR; age estimation; DNA methylation .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 1.Introduction DNA methylation age estimation is based on the dynamic changes of DNA methylation patterns throughout the lifespan. By comparing an individual's actual age with the level of DNA methylation in their genome, predictive models can be established to accurately infer age. DNA methylation age estimation holds broad applications and scientific significance. It can assess physiological and health conditions, explore associations between DNA methylation and aging, disease risks, and provide references for personalized medicine and health management. For instance, by comparing DNAm age with chronological age, potential health risks can be identified early, enabling appropriate preventive measures. Analyzing differences in DNAm age among different groups can reveal the impact of environmental factors (such as lifestyle, diet, and pollution) on biological age, thus informing health policies and interventions. In forensic practice, DNA methylation age estimation offers a novel approach for forensic medicine and individual identification. Analyzing DNA methylation levels to deduce biological age provides crucial information in legal contexts. Firstly, DNAm age can aid in criminal investigations by accurately determining the age of suspects using DNA methylation analysis of extracted samples. Secondly, it facilitates age determination in cases involving unidentified individuals, such as .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint forensic examinations of bodies or child abduction cases. Thirdly, DNAm age can assist in defining criminal responsibility age, which varies between minors and adults in many legal systems. While DNA methylation holds significant potential in forensic applications, rigorous research and practical experience are essential to ensure its scientific validity, stability, and accuracy in serving judicial practices objectively and fairly. With the advancing field of epigenetics, the focus in human characterization studies is shifting towards single nucleotide polymorphisms (SNPs), alongside traditional markers like restriction fragment length polymorphisms (RFLP) and short tandem repeats (STR). DNA methylation, as a crucial epigenetic marker, complements these markers and plays a pivotal role in age estimation models based on its dynamic changes across the lifespan. Traditional methylation models typically employ linear regression requiring multiple markers and samples, thereby increasing complexity and costs. This study explores SVR modeling principles to predict age, utilizing fewer but stronger linear markers, aiming to establish a highly stable and accurate blood DNA methylation age estimation model. Blood-based DNA methylation age estimation has been a prominent research area. Previous studies have utilized technologies like Illumina Infinium 450K chips and EpiTYPER systems to establish age prediction .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint models based on various CpG sites, demonstrating different levels of mean absolute deviation (MAD). The evolution of these models continues with improvements in SVR models, aiming for robustness and compatibility in age prediction from biological samples. In this study, a comparative analysis between multiple linear regression (MLR) and SVR models using sample data suggests that SVR models provide higher repeatability, better fit, and greater compatibility for developing accurate age prediction models based on blood DNA. Through literature review and analysis, this research aims to select CpG sites common in saliva and blood DNA, analyze their methylation levels using Illumina 850K chip technology, and establish a SVR regression model for blood-based DNA methylation age estimation. 2. Methods 2.1 Sample Collection This study utilized Illumina 850K chip data from 80 unrelated individual blood samples collected and tested by the Ministry of Public Security Appraisal Center. All samples were collected with informed consent from the individuals and approved by the ethics committee of the Ministry of Public Security Appraisal Center. The age of blood samples ranged from 20 to 59 years, with an average age of 39.08 years, as shown in Tab 1. .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 2.2 Selection of CpG Sites Based on literature review and laboratory data from the Ministry of Public Security Appraisal Center, 5 CpG sites potentially associated with age were initially selected as candidate sites for model training and research. These sites were numbered accordingly (as shown in Tab 2). Tab.1 Blood sample distribution information Age groups Sample size Average age 20 ~ 29 20 23.55 30 ~ 39 20 34.75 40 ~ 49 20 44.45 50 ~ 59 20 53.55 Tab.2 Initial screening site number information Identification SITE CpG1 CG21572722 CpG2 CG19283806 CpG3 CG03607117 CpG4 CG13552692 CpG5 CG14692377 CpG6 CG25410668 CpG7 CG04875128 CpG8 CG13552692 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 2.3 Pearson Correlation Analysis Pearson correlation analysis is a statistical method to assess the linear relationship and direction between two variables. In this experiment, the "cor()" function in RStudio was used to analyze data from 80 samples across 5 candidate CpG sites detected by the 850K chip. This analysis evaluated the correlation between methylation levels at each candidate site and the corresponding age of the samples, yielding correlation coefficients (denoted as "r") and regression scatter plots. 2.4 Selection and Optimization of Site Data To meet the significant correlation requirement between variables using Support Vector Machine (SVM), CpG sites with |r| > 0.5 from Pearson correlation analysis were selected for subsequent SVR regression model training and construction. Outlier samples were identified and excluded during data optimization to enhance prediction accuracy of the model. Normalization of selected sample data based on MATLAB R2022a's "mapminmax('apply')" function was performed to mitigate the impact of differing magnitude levels on model performance and computation speed. The final prediction values were reverse-normalized using "mapminmax('reverse')" to obtain age predictions in standard numerical ranges. 2.5 Construction of Age prediction Model Samples, after selection, were randomly shuffled with 80% used for .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint training and 20% for testing the model. The SVR regression model was built using MATLAB R2022a and the libsvm-3.34 toolkit, employing grid search and cross-validation to optimize parameters C (regularization parameter) and γ (gamma). The "svmtrain()" function established the model for predicting age from methylation levels at CpG sites, while "svmpredict()" simulated age predictions based on methylation levels in the test set. Visualization and mathematical analysis of model performance were conducted through plotting and calculation of relevant error metrics. 2.6 Analysis of Model Accuracy The accuracy of the age inference model was analyzed across four age groups (20-29, 30-39, 40-49, 50-59 years) using 76 selected blood samples. Model performance indicators such as Mean Absolute Deviation (MAD) and accuracy were computed, with an error margin set at ±5 years considered acceptable. Tab.3 R-value corresponding to candidate sites Site R CPG1 0.46 CPG2 -0.54 CPG3 0.61 CPG4 -0.52 CPG5 0.50 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 3. Results 3.1 Pearson Correlation Analysis Based on Illumina 850K chip data from 80 blood samples, DNA methylation levels at 8 CpG sites were correlated with sample age, as depicted in scatter plots in Fig.1. The calculated Pearson correlation coefficients are presented in Tab. 3. Among the 8 candidate sites, CpG2, CpG3, CpG4, CpG5, and CpG8 showed |r| values above 0.5, indicating a strong positive correlation between methylation levels and sample age. CpG8, while more correlated than CpG5, exhibited more outliers, which were unfavorable for SVR model construction. Therefore, CpG2, CpG3, CpG4, and CpG5 were chosen for the SVR model. 3.2 Site Selection and Data Processing Based on Pearson correlation analysis, CpG2, CpG3, CpG4, CpG5, and CpG8 showed |r| values above 0.5, indicating a strong positive correlation between methylation levels and sample age. However, CpG8's numerous outliers led to its exclusion from the study. Outlier samples were identified and excluded from the training set, ensuring robust model performance. Details are provided in the Tab.4. CPG6 0.37 CPG7 0.47 CPG8 -0.52 Tab.4 Abandoned blood sample information .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint Sample number Age Age stage DZ-484 31 20~29 JXH061 42 40~49 JXH080 45 40~49 JXH138 45 40~49 JXS001 46 40~49 L00341 50 40~49 L00490 51 50~59 L00398 57 50~59 L00445 L00445 57 57 50~59 50~59 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint Fig.1 Regression scatter plot of DNA methylation levels at 5 CpG sites in blood samples and sample age .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 3.3 Construction of Blood DNA Methylation Age prediction Model Using data from 70 selected samples, a model was built based on CpG2, CpG3, CpG4, and CpG5 sites. A mathematical formula for age prediction was derived using SVR regression, considering methylation levels at these sites. F{x} = [x1, x2, x3，x4]，F{ω} = [ω1，ω2，ω3，ω4] yp = F{ω} · F{x} + b Yp = yp × （N — M）+ M The model was trained and tested using MATLAB R2022a, yielding an MAE of 2.77 years and an accuracy of approximately 91.56% for the test set. The model's performance was validated through cross-validation, ensuring its stability and accuracy. The selection of training model parameters and formula variable values is shown in Tab. 5. Yp = 40 × [ F{ω} · F{x} -0.40263 ] + 20 Tab.5 Training model parameters and formula variable values Parameter (variable) Value Meaning .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint According to the calculation formula of relevant indicators, the training set R2 of the established SVR regression model is 0.90808, and the testing set R2 is 0.88529. Therefore, the age inference model t 2 RBF kernel C 1 Regularization parameter for regression models g 8 Parameters of the RBF kernel: gamma s 3 Train a model with epsilon-Support Vector Regression p 0.01 The epsilon value in the loss function ω1 -0.9350 Estimate CpG2 ω2 0.8352 Estimate CpG3 ω3 -0.8754 Estimate CpG4 ω4 0.7378 Estimate CpG5 b -0.4026 3 bias term in a regression model M 20 setting the left boundary of the sample interval N 60 setting the right boundary of the sample interva .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint constructed based on the CpG2-CpG3-CpG4-CpG5 locus combination can achieve an accuracy of about 89% for age prediction inference values, with a MAE value of 1.58 years for predicting age in the training set and actual age; The MAE value between the predicted age and the actual age in the test set is 2.57 years old. In the SVR regression model constructed this time, the fitting line graph between the predicted age of the training set and the true age is shown in Fig. 2, and the fitting line graph between the predicted age of the testing set and the true age is shown in Fig. 3. Fig.2 Fitting line graph between predicted age and actual age in the training set .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 图 3 测试集预测年龄与真实年龄的拟合折线图 Fig.3 Fitting line graph of predicted age and actual age in the test set 3.4 Accuracy Analysis across Different Age Groups The model's accuracy was assessed across four age groups using 60 training samples and 31 test samples. Performance metric ‘Accuracy’ are presented in Tab. 6 and Tab. 7 respectively. Despite decreasing Accuracy values with age group, indicating reduced predictive capability for older ages, the model demonstrated robust accuracy within ±5 years. Tab. 6: Accuracy metrics for age groups in the training set Age groups Accuracy Number of samples 20-29 94.1% 17 30-39 88.9% 18 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint 4. Discussion Recent studies on DNA methylation age inference have increasingly focused on using a small number of highly linear CpG sites for model construction. In this study, CpG2, CpG3, CpG4, and CpG5 showed optimal correlation (r-values) with sample age, facilitating SVR model construction. Through iterative training and refinement, a robust SVR mathematical model was established, achieving a 89.28% accuracy rate with a MAE of 2.77 years for age prediction. The model exhibited high stability and avoided overfitting, maintaining a self-check accuracy rate of 90%. However, as noted, MAD values increased with older age groups, suggesting a decline in predictive accuracy for older individuals. 40-49 84.6% 13 50-59 83.3% 12 Tab. 7: Accuracy metrics for age groups in the test set Age groups Accuracy Number of samples 20-29 100% 9 30-39 100% 8 40-49 75% 7 50-59 57.1% 7 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint Conflicts of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint

[1] Fraga MF, Ballestar E, Paz MF, Ropero S, Setien F, Ballestar ML, Heine-Suñer D, Cigudosa JC, Urioste M, Benitez J, Boix-Chornet M, Sanchez-Aguilera A, Ling C, Carlsson E, Poulsen P, Vaag A, Stephan Z, Spector TD, Wu YZ, Plass C, Esteller M. Epigenetic differences arise during the lifetime of monozygotic twins[J]. Proc Natl Acad Sci U S A, 2005, 102(30):10604-10609. [2] Christensen BC, Houseman EA, Marsit CJ, Zheng S, Wrensch MR, Wiemels JL, Nelson HH, Karagas MR, Padbury JF, Bueno R, Sugarbaker DJ, Yeh RF, Wiencke JK, Kelsey KT. Aging and environmental exposures alter tissue-specific DNA methylation dependent upon CpG island context[J]. PLoS Genet, 2009, 5(8):e1000602. [3]HORVATH S. Erratum to DNA methylation age of human tissues and cell types[J]. Genome Biol， 2015，16：96. [4]Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B, Bibikova M, Fan JB,Gao Y, Deconde R, Chen M, Rajapakse I, Friend S, Ideker T, Zhang K. Genome-wide methylation profiles reveal quantitative views of human aging rates[J]. Mol Cell, 2013,49(2):359-367. [5] Zbie ć-Piekarska R, Spólnicka M, Kupiec T, Parys-Proszek A, Makowska Ż, Pa łeczka A, Kucharczyk K, Płoski R, Branicki W. Development of a forensically useful age prediction method based on DNA methylation analysis[J]. Forensic Sci Int Genet,2015 ,17:173-179. [6] Horvath S. DNA methylation age of human tissues and cell types[J]. Genome Biology, 2013, 14(10):R115. [7] Levine ME, Lu AT, Quach A, Chen BH, Assimes TL, Bandinelli S, Hou L, Baccarelli AA, Stewart JD, Li Y, Whitsel EA, Wilson JG, Reiner AP, Aviv A, Lohman K, Liu Y, Ferrucci L, Horvath S. An epigenetic biomarker of aging for lifespan and healthspan[J]. Aging (Albany NY). 2018,10(4):573-591. [8]LIU C，MARIONI R E，HEDMAN A K ，et al. A DNA methylation biomarker of alcohol consumption[J]. Mol Psychiatry，2018，23（2）：422-433. [9]Hao T, Guo J, Liu J, Wang J, Liu Z, Cheng X, Li J, Ren J, Li Z, Yan J, Zhang G. Predicting human age by detecting DNA methylation status in hair[J]. Electrophoresis, 2021, 42(11): 1255-1261. [10] Belsky DW, Moffitt TE, Cohen AA, Corcoran DL, Levine ME, Prinz JA, Schaefer J, Sugden K, Williams B, Poulton R, Caspi A. Eleven telomere, epigenetic clock, and biomarker-composite quantifications of biological aging: do they measure the same thing?[J]. American Journal of Epidemiology, 2018, 187(6):1220-1230. [11]Zhang, Y., Wang, Y., & Shen, H. B. . DeepAge: a novel deep learning model for accurate age prediction based on DNA methylation. BMC genomics,2021, 22(1), 1-14. [12]Horvath, S., & Raj, K.DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nature Reviews Genetics, 2021, 22(3), 173-185. [13] Smith BH, Campbell A, Linksted P, Fitzpatrick B, Jackson C, Kerr SM, Deary IJ, Macintyre DJ, Campbell H, McGilchrist M, Hocking LJ, Wisely L, Ford I, Lindsay RS, Morton R, Palmer CN. Cohort profile: Generation Scotland: Scottish Family Health Study (GS: SFHS). The study, its participants and their potential for genetic research on health and illness[J]. International Journal of Epidemiology, 2013, 42(3):689-700. .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint [14]Wang, L., Zhao, Y., Xu, Y., Xu, Z., Yao, Y., Huang, L., ... & Liu, Y. Effects of environmental exposures on DNA methylation-based age prediction: A study in a Chinese population. Environmental Pollution, 2023 292, 118419. [15]Li, Z., Zhang, W., Li, J., Zhou, L., & Wang, H.Cross-tissue DNA methylation age prediction based on single-cell sequencing data. Bioinformatics, 2024,40(1), 122-130. [16]Zhu, Y., Huang, T., & Zhou, X. Identification of DNA methylation markers associated with longevity using artificial intelligence. Aging Cell,2024,23(3), e13500. [17]Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B, Bibikova M, Fan JB,Gao Y, Deconde R, Chen M, Rajapakse I, Friend S, Ideker T, Zhang K. Genome-wide methylation profiles reveal quantitative views of human aging rates[J]. Mol Cell, 2013,49(2):359-367. [18] Zbieć-Piekarska R, Spólnicka M, Kupiec T, Parys-Proszek A, Makowska Ż, Pałeczka A, Kucharczyk K, Płoski R, Branicki W. Development of a forensically useful age prediction method based on DNA methylation analysis[J]. Forensic Sci Int Genet,2015 ,17:173-179. [19]Xu C, Qu H, Wang G, Xie B, Shi Y, Yang Y, Zhao Z, Hu L, Fang X, Yan J, Feng L. A novel strategy for forensic age prediction by DNA methylation and support vector regression model[J]. Scientific reports, 2015, 5(1): 1-10. [20] XU C ， QU H ， WANG G ， et al. A novel strategy for forensic age prediction by DNA methylation and support vector regression model[J]. Sci Rep，2015，5：17788. [21]Feng L, Peng F, Li S, Jiang L, Sun H, Ji A, Zeng C, Li C, Liu F. Systematic feature selection improves accuracy of methylation-based forensic age estimation in Han Chinese males[J]. Forensic Science International: Genetics, 2018, 35: 38-45. [22] Pan C, Yi S, Xiao C, Huang Y, Chen X, Huang D. The evaluation of seven age-related CpGs for forensic purpose in blood from Chinese Han population[J]. Forensic Sci Int Genet,2020 ,46:102251. [23]Smith, Z. D., & Meissner, A. DNA methylation: roles in mammalian development. Nature Reviews Genetics,2013，14(3), 204–220. [24]Jones, P. A., & Takai, D.The r ole of DNA methylation in mammalian epigenetics. Science, 2001,293(5532), 1068–1070. [25]Lu, A. T., Quach, A., Wilson, J. G., Reiner, A. P., Aviv, A., Raj, K., ... & Horvath, S. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging,2020,12(10), 9711-9728. [26]Knight, A. K., Craig, J. M., Theda, C., Bækvad-Hansen, M., Bybjerg-Grauholm, J., Hansen, C. S., ... & Hollegaard, M. V. The epigenetic clock during human aging and the aging of stem cells and their niches. Nature Aging,2022, 2(4), 254-271. [27]Gao, X., Jia, M., Zhang, Y., Liu, H., Li, X., & Yang, X. Dynamic DNA methylation patterns in the nervous system and their association with neurodegenerative diseases. Neuroscience & Biobehavioral Reviews,2024,128, 82-94. [28]Zhang, Z., Li, J., Yu, Y., & Li, D.. Application of DNA methylation in forensic science. Forensic Sciences Research, 2020,5(4), 399-405. [29]Ballard, D., Giese, R. W., & Martello, S. DNA methylation age estimation in bloodstains: A powerful tool for forensic science. Forensic Science International: Genetics, 2020,48, 102340. [30]Freire-Aradas, A., Phillips, C., Lareu, M. V., & Carracedo, Á.Forensic individual age estimation with DNA: From initial approaches to methylation tests. Forensic Science Reviews,2016, 28(2), 111-126. .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted October 29, 2024. ; https://doi.org/10.1101/2024.10.25.620181doi: bioRxiv preprint