Environmental and Maternal Imprints on Infant Gut Metabolic Programming

Environmental and Maternal Imprints on Infant Gut Metabolic Programming | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Environmental and Maternal Imprints on Infant Gut Metabolic Programming View ORCID Profile Kine Eide Kvitne , View ORCID Profile Celeste Allaband , Jennifer C. Onuora , Daniela Perry , View ORCID Profile Simone Zuffa , Lucas Patel , View ORCID Profile Vincent Charron-Lamoureux , View ORCID Profile Ipsita Mohanty , View ORCID Profile Kristija Sejane , View ORCID Profile Abubaker Patan , Abdullah Al Mahmud , Tahmeed Ahmed , View ORCID Profile Diego G. Bassani , Antonio González , View ORCID Profile Davidson H. Hamer , Rashidul Haque , Benjamin Ho , Md Iqbal Hossain , Mohammad Shahidul Islam , View ORCID Profile Daniel McDonald , Lisa G. Pell , Huma Qamar , View ORCID Profile Daniel E. Roth , Samir Saha , View ORCID Profile Prakesh S. Shah , Md Muniruzzaman Siddiqui , Shafiqul Alam Sarker , Shamima Sultana , Sydney Thomas , Lindsey A. Burnett , Shirley M. Tsunoda , Lars Bode , Pieter C. Dorrestein , Rob Knight doi: https://doi.org/10.1101/2025.07.24.666662 Kine Eide Kvitne 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Kine Eide Kvitne Celeste Allaband 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Celeste Allaband Jennifer C. Onuora 3 Centre for Global Child Health, The Hospital for Sick Children , Toronto, Ontario, Canada 4 Department of Nutritional Sciences, Faculty of Medicine, University of Toronto , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniela Perry 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 5 Bioinformatics and Systems Biology Program, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Simone Zuffa 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA 6 Collaborative Mass Spectrometry Innovation Center, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Simone Zuffa Lucas Patel 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 5 Bioinformatics and Systems Biology Program, University of California San Diego , La Jolla, CA, USA 7 Medical Scientist Training Program, University of California , San Diego, La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Vincent Charron-Lamoureux 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA 6 Collaborative Mass Spectrometry Innovation Center, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vincent Charron-Lamoureux Ipsita Mohanty 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA 6 Collaborative Mass Spectrometry Innovation Center, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Ipsita Mohanty Kristija Sejane 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 8 Larsson-Rosenquist Foundation Mother-Milk-Infant Center of Research Excellence (MOMI CORE), University of California San Diego , La Jolla, CA, USA 9 Human Milk Institute (HMI), University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Kristija Sejane Abubaker Patan 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA 6 Collaborative Mass Spectrometry Innovation Center, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Abubaker Patan Abdullah Al Mahmud 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh 11 Shield Pharmaceuticals Corp. , Ronkonkoma, NY, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tahmeed Ahmed 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Diego G. Bassani 3 Centre for Global Child Health, The Hospital for Sick Children , Toronto, Ontario, Canada 12 Child Health Evaluative Sciences, The Hospital for Sick Children , Toronto, Ontario, Canada 13 Dalla Lana School of Public Health, University of Toronto , Toronto, Ontario, Canada 17 Department of Paediatrics, Faculty of Medicine, University of Toronto , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Diego G. Bassani Antonio González 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Davidson H. Hamer 14 Department of Global Health, Boston University School of Public Health , Boston, MA, USA 15 Department of Medicine, Boston University Chobanian & Avedisian School of Medicine , Boston, MA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Davidson H. Hamer Rashidul Haque 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Benjamin Ho 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Md Iqbal Hossain 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mohammad Shahidul Islam 16 Child Health Research Foundation , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel McDonald 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Daniel McDonald Lisa G. Pell 3 Centre for Global Child Health, The Hospital for Sick Children , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Huma Qamar 3 Centre for Global Child Health, The Hospital for Sick Children , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel E. Roth 3 Centre for Global Child Health, The Hospital for Sick Children , Toronto, Ontario, Canada 4 Department of Nutritional Sciences, Faculty of Medicine, University of Toronto , Toronto, Ontario, Canada 12 Child Health Evaluative Sciences, The Hospital for Sick Children , Toronto, Ontario, Canada 13 Dalla Lana School of Public Health, University of Toronto , Toronto, Ontario, Canada 17 Department of Paediatrics, Faculty of Medicine, University of Toronto , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Daniel E. Roth Samir Saha 16 Child Health Research Foundation , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Prakesh S. Shah 17 Department of Paediatrics, Faculty of Medicine, University of Toronto , Toronto, Ontario, Canada 18 Department of Pediatrics, Mt. Sinai Hospital , Toronto, Ontario, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Prakesh S. Shah Md Muniruzzaman Siddiqui 19 Mohammadpur Fertility Services and Training Centre , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Shafiqul Alam Sarker 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Shamima Sultana 10 Nutrition Research Division, International Centre for Diarrhoeal Disease Research , Bangladesh , Dhaka, Bangladesh Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sydney Thomas 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lindsey A. Burnett 20 Department of Obstetrics Gynecology and Reproductive Sciences, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Shirley M. Tsunoda 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lars Bode 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 8 Larsson-Rosenquist Foundation Mother-Milk-Infant Center of Research Excellence (MOMI CORE), University of California San Diego , La Jolla, CA, USA 9 Human Milk Institute (HMI), University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Pieter C. Dorrestein 1 Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego , La Jolla, CA, USA 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 6 Collaborative Mass Spectrometry Innovation Center, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: rknight{at}health.ucsd.edu pdorrestein{at}health.ucsd.edu Rob Knight 2 Department of Pediatrics, University of California San Diego , La Jolla, CA, USA 21 Center for Microbiome Innovation, University of California San Diego , La Jolla, CA, USA 22 Department of Computer Science and Engineering, University of California San Diego , La Jolla, CA, USA 23 Shu Chien-Gene Lay Department of Bioengineering, University of California San Diego , La Jolla, CA, USA 24 Halıcıoğlu Data Science Institute, University of California San Diego , La Jolla, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: rknight{at}health.ucsd.edu pdorrestein{at}health.ucsd.edu Abstract Full Text Info/History Metrics Supplementary material Preview PDF ABSTRACT Early life is a critical period for immune and metabolic programming, but developmental patterns remain underexplored in populations from low- and middle-income countries. Here, we profiled the microbiome and metabolome of 55 Bangladeshi mother-infant dyads over the first six months of life. Importantly, we observed an increase in microbially-derived bile amidates and N -acyl lipids with age in conjunction with reads matching the bile salt hydrolase/transferase ( bsh ) gene. While microbial source tracking confirmed maternal fecal seeding, a substantial environmental contribution was also highlighted. Differences in infant fecal metabolic profiles were associated with delivery mode, maternal milk composition, household assets, and household-level water treatment. C-section delivery and untreated drinking water were linked to transient metabolic differences, including increases in bile amidates, N -acyl lipids, and other host-microbe co-metabolic products, including acylcarnitines. Multi-omics analysis revealed specific microbial-metabolite relationships, highlighting how early environmental and maternal living circumstances shape metabolic gut programming through the microbiome. MAIN The first months after birth represent a crucial window for infant development, laying the groundwork for long-term health. During this period, the establishment and maturation of the gut microbiome play a key role in metabolic programming and training of the infant immune system 1 , 2 . Although these processes are known to be shaped by a complex interplay between maternal and environmental factors 3 , 4 , our current knowledge largely originates from studies in high-income countries 5 . Thus, early-life gut microbiome composition and function in low- and middle-income countries (LMICs) largely remain underexplored. Delivery mode is one of the earliest determinants of newborn microbiota colonization 6 , 7 . While vaginally born infants are seeded by maternal fecal and vaginal microbiota, infants born via Cesarean section (C-section) are primarily colonized by skin associated microorganisms, such as Staphylococcus and Corynebacterium 6 , 8 . C-section births have increased by nearly 30% over the last decade in many LMICs, including Bangladesh 9 . After birth, challenges with early initiation and duration of breastfeeding 10 – 12 can limit infant exposure to bioactive human milk oligosaccharides (HMOs). These molecules function as prebiotics, fostering the growth of the infant gut microbiota and influencing health outcomes 13 , 14 . For example, α1-2-fucosylated HMOs, like 2’-fucosyllactose (2’FL), produced almost exclusively by mothers with a functional fucosyltransferase 2 (FUT2) gene 15 (Secretors) can promote early-life growth of beneficial microorganisms, such as Bifidobacterium 16 , 17 . In contrast, the abundance of Bifidobacterium species is negatively impacted by antibiotic treatment in early infancy 18 , along with overall changes to microbiome diversity and composition 7 , 19 , 20 . Antibiotic usage is considerably more common in many LMICs 21 – 23 . In addition, variable household hygiene standards and sanitation practices 24 which are generally connected to wealth may further influence the microbiome development through different exposure to pathobionts or molecular exposures. Since infant microbiomes do not function in isolation, integrating microbiome and metabolomics data through multi-omics approaches is essential to understand how early-life exposures and living conditions shape gut metabolic programming. However, such efforts have been limited by the technical difficulty of acquiring rich, longitudinal multi-omics data, particularly in LMIC settings. Here, we address this gap through an observational, longitudinal study of 55 mother-infant dyads in Bangladesh, integrating metagenomics, untargeted metabolomics, and HMO profiling across the first six months of life. Our study demonstrates a rapid increase in microbially-derived metabolites, including bile amidates and N -acyl lipids, in response to the maturing gut microbiome. We also show that delivery mode, maternal secretor status, and household drinking water treatment are associated with different infant fecal metabolomic profiles. Finally, multi-omics integration reveals specific microbial-metabolite relationships, highlighting the role of the gut microbiome as a key mediator of early metabolic programming. RESULTS We enrolled 55 mother-infant dyads from two urban hospitals in Dhaka, Bangladesh, and sampled them at multiple timepoints from birth to six months postnatal age ( Fig. 1a , detailed flow chart in Extended Data Fig. 1a) . Infants between 0-1 days old and orally feeding at the time of screening were eligible for inclusion. We collected stool samples from both infants and mothers, human milk and vaginal swabs from mothers, and blood, skin (elbow), and oral cavity (tongue) samples from infants. Untargeted liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) metabolomics was used to analyze infant stools, plasma, and maternal milk, while shotgun metagenomic sequencing was performed on all sample types except plasma. A total of 1,609 metagenomics samples and 530 metabolomics samples were processed. Download figure Open in new tab Fig. 1: Rapid shifts in key signalling metabolites during infant fecal metabolic gut programming. a, Experimental overview. Longitudinal samples were collected from 55 mother-infant dyads in Dhaka, Bangladesh. Infant samples included stool, oral (tongue), skin (elbow), and blood; maternal samples included stool, milk, and a vaginal swab (see Methods for details). Open circles indicate random collection time within the first 60 days. n reflects the initial number of samples per biospecimen, prior to quality control. b, UpSet plot of metabolic features across infant fecal (n=377) and plasma (n=34) samples, and maternal milk (n=119) samples. The left panel shows features unique to or shared between sample types, while the right panel shows chemical classes and pathways predictions based on NPC classification from CANOPUS (probability > 0.5). c, Principal Component Analysis (PCA) of maternal milk showing significant development over time (Permutational Multivariate Analysis of Variance (PERMANOVA): p < 0.001). d, Boxplot showing the natural log peak areas of key discriminant features across lactation stages: colostrum ( 15 days). Statistical significance was determined by the Wilcoxon rank-sum test, with p values adjusted for multiple comparisons using the Benjamini-Hochberg (BH) method. Repeated measures were avoided by retaining only the highest value per subject within each lactation stage. e, PCA of fecal metabolic profiles across early infancy (0-6 months). Differences in the metabolic profiles were associated with infant age, delivery mode, maternal secretor status, and drinking water treatment (PERMANOVA: all p < 0.05). f, Longitudinal abundances of free di- and tri-hydroxylated bile acids in infant feces over the first 6 months of life. g, Abundance trajectories of taurine (tau), glycine (gly), and amino acid (AA) conjugated bile acids, as well as oxo bile acids in the same period. h, Scatter plot showing the association between the natural log of bile salt hydrolase/transferase ( bsh) gene read count matches and the ratio of amino acid to taurine conjugated bile acids using Pearson correlation for significance (linear mixed effect model (LME): β = 0.25, p < 0.001). A linear regression line with 95% CI is shown, with points colored by infant age. i, Longitudinal abundance trajectories of N -acyl lipids with a C2:0 fatty acid chain (left panel) and a C3:0 fatty acid chain (right panel) amidated to different head groups. j, Longitudinal abundances of MS/MS spectral matches to various short-chain acylcarnitines (left panel) and long-chain acylcarnitines (right panel) in infant feces. For panels f, g, i, j , β and p values were derived from LME with subject as a random effect, and scatter plot smoothing (‘LOESS’) was used for visualization purposes. All boxplots show the first (lower), median, and third (upper) quartiles, with whiskers 1.5 times the interquartile range. Characteristics of the study cohort are detailed in Table 1 . In short, 40% of infants were male and infants had an average birth weight of 2,924 ± 428 g (mean ± standard deviation). All mothers received postpartum antibiotics, and, in line with the pre-specified enrollment protocol, there was an even distribution between vaginal and C-section deliveries. Exclusive breastfeeding rates declined from 62% at one month to 42% at 3 months to a steep drop of 16% at 6 months. Based on 2’-fucosyllactose (2’FL) levels, 67% of mothers were classified as secretors. Because variable household hygiene standards in LMICs may also influence microbiome composition and function, we also gathered data on variables such as drinking water practices. Importantly, 67% of households reported using at least one method to treat their drinking water (boiling and/or filtering). View this table: View inline View popup Download powerpoint Table 1. Characteristics of the study cohort. Numbers are given as mean ± SD or count (%). Infant Age Shapes Early Metabolic Trajectories in Maternal Milk and Infant Feces To characterize early-life metabolic trajectories in the mother-infant dyads, we profiled infant feces (n=377), infant plasma (n=34), and maternal milk (n=119) using untargeted LC-MS/MS metabolomics. This resulted in the detection of 13,871 metabolic features, of which 1,735 were shared across all three biospecimens ( Fig. 1b ) . As expected, each sample type displayed a distinct metabolic composition. Maternal milk was rich in amino acids, saccharides, and triacylglycerols, while infant feces had a broader range of metabolite classes, including amino acids, small peptides, fatty acids, N -acyl amines, steroids, sphingolipids, and alkaloids. Maternal milk composition changed over time (permutational multivariate analysis of variance (PERMANOVA): R 2 = 0.04, F-statistic = 5.1, p < 0.001; Fig. 1c ), consistent with knowledge that nutritional needs change as the infant grows 26 . To identify metabolic changes across lactation stages, milk samples were grouped into colostrum ( 15 days), and analysed using a supervised Partial Least Squares Discriminant Analysis (PLS-DA) model (classification error rate (CER) = 0.18; Extended Data Fig. 2a ). Features with a variable importance in projection (VIP) > 1 were extracted ( Supplementary Table 1 ). Colostrum was enriched in amino acids, small peptides, and short-chain acylcarnitines, including acetylcarnitine (C2:0) and propionylcarnitine (C3:0) ( Fig. 1d , Extended Data Fig. 2b ). Here, C3:0 refers to a fatty acid chain with three carbons and no double bonds 27 . In contrast, mature milk contained higher levels of carbohydrate sources, such as mono- and di-saccharides, as well as glutamine, glutamate and butyrylcarnitine (C4:0), consistent with previous findings 26 , 28 . We also observed a clear development of the infant fecal metabolome with age (PERMANOVA: R 2 = 0.012, F-statistic = 4.5, p < 0.001; Fig. 1e ). Bile acid metabolism, critical for infant development 29 , changed during the first weeks of life. Using recently developed libraries 30 – 32 , we show that the fecal abundance of free di- and tri-hydroxylated bile acids increased over the first weeks of life (linear mixed effects model (LME): di: β = 0.035, p = 9.4×10 -10 ; tri: β = 0.015, p = 7.9×10 -9 ; Fig. 1f ). Host-derived taurine (LME: β = -0.018, p = 1.9×10 -6 ) and host-microbially derived glycine (LME: β = -0.015, p = 0.034) conjugates decreased in the same period ( Fig. 1g ). In contrast, microbially modified bile acids, including various amino acid conjugates (LME: β = 0.026, p = 8.9×10 -5 ) and oxo bile acids (LME: β = 0.025, p = 4.0×10 -6 ) increased ( Fig. 1g ). This opposite trend suggests a growing capacity of the gut microbiome to modify host-derived bile acids as it matures. To investigate this more mechanistically, we looked at the microbial bile salt hydrolase/transferase ( bsh ) gene, which deconjugates taurine and glycine conjugates through hydrolysis of the amide bond but also are capable of conjugating bile acids with amines 33 , 34 . We observed a positive correlation between bsh gene levels and the ratio of amino acid to taurine conjugated bile acids (Pearson correlation: r = 0.49, p = 7.5e-10 ; Fig. 1h ). Notably, both the abundance of the bsh gene family (LME: β = 0.057, p < 0.001; Extended Data Fig. 2c ) and this bile acid ratio (LME: β = 0.074, p < 0.001; Extended Data Fig. 2d ) increased with age, linking the maturation of the gut microbiome to the observed shifts in the bile acid pool. Another group of signalling molecules that underwent changes in early life were the N -acyl lipids, metabolites consisting of a fatty acyl chain amidated with an amino containing group 27 . Specifically, we observed an increase in the levels of short-chain N -acyl lipids, including acetate (C2:0) and propionate (C3:0), across different head groups (LME: all p < 0.01; Fig. 1i ). Finally, we observed distinct developmental trajectories for the acylcarnitines based on their chain length ( Fig. 1j ). These compounds are possibly originating from host-microbial co-metabolism where microbially-produced short-chain fatty acids 35 are attached to carnitine. Long-chain (C16-C18) acylcarnitines increased during the first month of life (LME: all p 0.05). This is likely due to maternal transfer by breastfeeding, which was supported by the detection of these short-chain acylcarnitines in human milk. These metabolic changes co-occurred with maturation of the infant gut microbiome. Using the phylogenetic robust center log ratio (rclr) metric phylo-RPCA 36 to assess beta diversity, we demonstrated that age influenced infant fecal microbiota composition (PERMANOVA: F-statistic = 18.3, p < 0.001; Fig. 2a ) . Since the difference was most pronounced between samples collected before and after one month (30 days) of age, that time point was chosen as the inflection point for subsequent categorical differential abundance analysis. Using ANCOM-BC2, we identified specific taxa driving maturation of the gut microbiome (p adj < 0.05; Fig. 2b for top/bottom 12 microorganisms). Early gut colonizers such as Enterococcus 37 decreased over time, while several Bifidobacterium species became more abundant after the first month of life. Using the genera identified by ANCOM-BC2 ( Fig. 2b , Extended Data Fig. 3a ), we calculated a natural log ratio of genera enriched after one month (numerator) to genera depleted before one month (denominator) ( Fig. 2c ). An approximately 3-log fold increase was observed in older compared to younger infants (Day 0 mean ratio = -1.05; Day 60 mean ratio = 2.14) with a positive trend across all major timepoints (LME: β = 0.026, p < 0.001). While Faith’s phylogenetic diversity (PD) for infant fecal samples did not change over the study period ( Extended Data Fig. 3b-c ), there appeared to be a trend of lower values for the first 60 days that then began to increase, potentially due to antibiotic exposure. Download figure Open in new tab Fig. 2: Sources driving the developing gut microbiome. a, phylo-RPCA EMPeror plot of fecal samples with more than 100,000 reads (n=219) colored by infant age bin. b, Top and bottom 12 differentially abundant infant fecal genera identified by ANCOM-BC2 (p adj 0.5). Full plot shown in Extended Data Fig. 3a . Enriched (yellow) or depleted (purple) after 30 days (later timepoints) are shown, black error bars indicate standard error. c, Scatterplot with lineplot of the natural log-ratio of the summed counts of enriched (numerator) and depleted (denominator) fecal genera found in b over time (LME: β = 0.026, p < 0.001); lineplot used binned timepoints. FEAST : Infant fecal samples as sinks. d, Stacked barplot of the mean percentage contribution of each maternal body site on infant fecal samples for matched (left) or all available (right) maternal samples. e, Stripplot/pointplot of the FEAST estimated percentage contribution of maternal fecal samples to matched (left) or all available (right) infant fecal samples. f, Stripplot/pointplot of the FEAST estimated percentage contribution of unknown elements to matched (left) or all available (right) infant fecal samples. Notation: ** = p < 0.01. Further, we observed that Faith’s PD alpha diversity decreased for infant skin samples (LME: p < 0.001; Extended Data Fig. 3d-e ) and increased for infant oral samples (LME: p < 0.001; Extended Data Fig. 3f-g ) over time. Maternal fecal samples also displayed significant changes over the course of the study (LME: p < 0.001), but primarily in the first weeks after giving birth ( Extended Data Fig. 3h ), likely related to postpartum antibiotic administration. As expected, there was a difference in Faith’s PD between maternal and infant fecal samples during the study (LME: p = 0.012; Extended Data Fig. 3i ). There was no difference in the adult vaginal microbiome at the time of birth between delivery modes ( Extended Data Fig. 3j ). Similarly, there were no changes in the human milk microbiome over time ( Extended Data Fig. 3k ). All samples are compared using two beta diversity metrics, phylo-RPCA and Unweighted Unifrac, which showed differences between all body sites (PERMANOVA: p < 0.05 for all; Extended Data Fig. 4a,b ). Finally, when looking at the phylo-RPCA beta diversity of infant samples, we see overlap between body sites at birth (Day 0) but by 6 months (Day 180) we see body sites are separated (PERMANOVA: p adj < 0.01 for all; F-statistic: feces-skin 265, skin-tongue 195, feces-tongue 7) as the microbes adapt to each niche ( Extended Data Fig. 4c,d) . Maternal and Environmental Sources Seed the Developing Gut Microbiome To explore sources of microbiota colonization, we performed source tracking using FEAST (Fast Expectation-mAximization microbial Source Tracking) 38 . When using both matched and all available maternal samples as sources, FEAST estimated that, of the available biospecimen types, maternal feces contributed the most to infant fecal microbiome development ( Fig. 2d ). Matched mother-infant samples had a much larger proportion of unknown, likely environmental, contributions (mean unknown proportions: early = 0.85, late = 0.94), than when using all available adult samples as possible sources of microbes (mean unknown proportions: early = 0.32, late = 0.37). This was further supported by the fact that there were no differences in phylo-RPCA beta diversity differences within and between related mother-infant pairs ( Extended Data Fig 4e,f ). There was no significant difference in the proportion of the infant gut microbiome attributed to matched or all available maternal vaginal or human milk samples when comparing early and later timepoints ( Extended Data Fig. 4g,h ), indicating that bacteria seeded at birth did appear to persist, although there was individual infant variation. There was also a decrease in the proportion of the infant gut microbiome attributed to matched maternal fecal samples between early and late timepoints (2MWW, p = 0.0068), but no significant differences in proportions over time when considering all available maternal fecal samples as possible sources ( Fig. 2e ). When non-matched maternal fecal samples were considered, they appeared to contribute proportionally more to the infant gut microbiome (approximately 60%) compared to when only matched maternal fecal samples were examined (approximately 10%) ( Fig. 2e ). This suggests that environmental sources, including other household members and non-maternal caretakers 39 , 40 , play a larger role in the ongoing development of the infant microbiome than previously understood. There was also a significant increase in the contributions of unknown sources over time on the infant fecal microbiome with matched samples (2MWW, p = 0.0042), but not when considering all available sources ( Fig. 2f ). In general, the infant gut microbiome development trajectory appears to mature towards an adult-like state, rather than converging towards the mother’s specific microbiome as previously suggested 6 , 41 , 42 . Together, this indicates that while maternal seeding is important, a large proportion of the infant gut microbiome is derived from increasing amounts of environmental sources over time, at least in this Bangladeshi cohort. Cesarean Delivery is Associated with Higher Levels of Microbially Derived Metabolites in Early Infancy While infant age was the primary variable shaping the fecal metabolome, PERMANOVA also identified other significant factors, including delivery mode ( Fig. 3a ). The model output for all tested variables is provided in Supplementary Table 2 . The natural log ratios of features associated with C-section (numerator) and vaginal birth (denominator) identified by PLS-DA (CER = 0.40; Extended Data Fig. 5a ) separated groups over time (LME: β = -0.71, p = 2.4×10 -5 ; Fig. 3b ), with the largest difference observed during the first month of life. Infants born via C-section were enriched in amino acids, dipeptides, acetylcarnitine (C2:0), and N-acetylneuraminic acid (Neu5Ac). Neu5Ac is the primary sialic acid component found within sialylated HMOs in human milk 15 , with roles in host-microbe interactions 43 and neurodevelopment 44 , 45 . Conversely, vaginally delivered infants showed enrichment of several microbially derived bile acids and saccharides ( Supplementary Table 1 ). Furthermore, different N -acyl lipids were enriched in C-section and vaginally born infants, varying in head group and chain length ( Supplementary Table 1 ). Surprisingly, the effect of delivery mode on infant gut microbiome composition was less pronounced, possibly due to the fact that all mothers received postpartum antibiotics, only reaching statistical significance at Day 7 (2MWW-HB, p = 0.006; Extended Data Fig. 5b ), but not over the entire study period (LME: p = 0.118). However, we did see a statistically significant effect of delivery mode on Faith’s phylogenetic alpha diversity over time (LME: β = 11.8, p = 0.029). Download figure Open in new tab Fig. 3: Delivery mode and water treatment are associated with infant fecal metabolic trajectories in early-life. a, PCA of infant fecal samples showing separation by delivery mode (PERMANOVA: p < 0.001). b , Longitudinal natural log ratios of the summed peak areas of the top 100 discriminating features for delivery mode in infant feces. LME was used to determine statistical significance between groups. Scatter plot smoothing (‘LOESS’) and removal of three outliers for visualization. c , Barplot of metabolic features significantly differing between C-section and vaginal birth (p adj < 0.10) in infant feces collected during days 10-30 after birth. For each infant, the sample with the highest log ratio within this interval was selected. d , Boxplot of the natural log ratios of amino acid to taurine conjugated bile acids by delivery mode at selected timepoints (Wilcoxon rank-sum test, day 14-28: p = 0.042, p adj = 0.13). e , Boxplot of bsh gene reads in infant feces by delivery mode (Wilcoxon rank-sum test, day 14-28: p = 0.07, p adj = 0.08). f , Trajectories for valine-C2:0 and 2-aminovaleric acid-C2:0 by delivery mode in infant feces, with smoothing for visualization. The difference between C-section and vaginal birth was assessed for significance using LME. g , PCA of infant fecal samples showing separation by household drinking water treatment (PERMANOVA: p = 0.020). h , Longitudinal natural log ratios of the summed peak areas of the top 100 discriminating features for treated vs. untreated drinking water. Smoothing added for visualization, and statistical significant difference between groups was assessed using LME. i, Molecular network of acylcarnitines associated with household drinking water treatment (untreated=brown, treated=dark green). Node size represents mean VIP score from three different PLS-DA models accounting for different purification methods, with larger nodes indicating higher VIP score. Only acylcarnitines that were significant in all three models (VIP > 1) and with the same directionality were included in the network. j, Boxplot showing differences in the natural log peak areas of various acylcarnitines enriched in the untreated water group at day 0-7. Statistical significance was determined by the Wilcoxon rank-sum test with BH adjusted p values. k, Differentially abundant infant fecal genera by water treatment identified by ANCOMB-BC2 (q 1). Enriched (grey) or depleted (brown) in treated water are shown, black errors indicate standard error. l, Scatterplot/lineplot of the natural log-ratio of the summed counts of enriched (numerator) and depleted (denominator) fecal genera found in k (LME: -0.058, p = 0.13); lineplot used binned timepoints. For b, f, and h, β and p values were derived from LME models with subject as random effect. All boxplots show the first (lower), median, and third (upper) quartiles, with whiskers 1.5 times the interquartile range. For boxplots, repeated measures were avoided by retaining only the highest value per individual within each time window. To further explore the divergence in metabolic trajectories based on delivery mode, we focused on days 10-30 after birth, the earliest point of group separation. Most discriminant features during this window were enriched in infants delivered via C-section and included bile acids, short-chain acylcarnitines, N -acyl lipids, amino acids, tryptophan derivatives, dipeptides, and lipid derivatives ( Fig. 3c ). Compounds such as indolelactic acid and dipeptides are known metabolic products of the gut microbiome 46 , 47 , often generated through microbial digestion of proteins into peptides and amino acids. In contrast, features enriched in vaginally delivered infants included milk-associated glycans ( Fig. 3c ). Given the observed differences in several bile acids, we further investigated the ratio of microbiome-derived amino acid to host-derived taurine conjugates and the levels of microbial bsh gene in infant stool. Interestingly, we observed a trend where both this ratio and bsh gene levels were transiently elevated in C-section compared to vaginally born infants between days 14-28, though these differences did not reach statistical significance after FDR correction (Wilcoxon rank-sum test, p adj = 0.13 and p adj = 0.08, respectively; Fig. 3d,e ). This may indicate a temporary enrichment of microorganisms with more bsh genes in C-section born infants the first weeks of life, however, this needs further investigation in a larger cohort. Further, we observed transiently elevated levels of valine-C2:0 (LME: β = -3.2, p = 1.9×10 -4 ) and 2-aminovaleric acid-C2:0 (LME: β = -2.3, p = 0.0042) in C-section vs. vaginally born infants ( Fig. 3f ). A similar pattern was observed for acetylcarnitine (C2:0), propionylcarnitine (C3:0), and hexanoylcarnitine (C6:0), with higher levels between days 14-28 (Wilcoxon rank-sum with BH correction, all p < 0.05; Extended Data Fig. 5c ). Although we observed only minor differences in the gut microbiome composition based on delivery mode, its metabolic output may differ, as suggested by the distinct profiles obtained for the microbially derived metabolites, including those involved in immune regulation pathways, such as bile acids 30 and N -acyl lipids 27 . This may contribute to differences in early-life immune maturation observed between delivery modes 8 , 48 . Indeed, early-life changes in bile amidates have recently been found to be predictive of islet autoimmune dysregulation in diabetic onset post-birth while N -acyl lipids controlled T-cell differentiation 49 . Household Drinking Water Treatment is Associated with Altered Carnitine Metabolism Consistent with the source tracking analyses showing a significant environmental contribution to the infant gut microbiome, both socioeconomic status (asset index, PERMANOVA: F-statistic = 1.5, p = 0.006; Extended Data Fig. 5d-f ) and reported household drinking water treatment practices (PERMANOVA: F-statistic = 1.4, p = 0.020; Fig. 3g ) were associated with different infant fecal metabolic profiles. We further explored the relationship to drinking water treatment, viewing it as an indirect proxy for household hygiene and sanitation practices, given previously reported association with the gut microbiome 50 , 51 . The natural log ratios of features associated with treated water (numerator) and untreated water (denominator) identified by PLS-DA (CER = 0.35; Extended Data Fig. 5g ) clearly separated groups throughout the study period (LME: β = -0.36, p = 1.6×10 -5 ; Fig. 3h ). To account for different purification methods (e.g. boiling vs. filtering), we built three separate PLS-DA models ( Extended Data Fig. 5h-i ). Despite that the overall model performances were modest (CER 1 and same directionality revealed 255 common metabolic features ( Supplementary Table 3 ). The most prominent finding was a consistent enrichment of multiple acylcarnitines in the untreated water group ( Fig. 3i ), compounds central to fatty acid metabolism, host mitochondrial function, and energy metabolism 52 . Interestingly, the difference in carnitine metabolism was most pronounced the first weeks of life ( Fig 3j ). Given that acylcarnitines are present in human milk, with higher levels of acetylcarnitine (C2:0) and propionylcarnitine (C3:0) in colostrum as shown previously, we investigated if water treatment also influenced the milk metabolome. Although not significant in the PERMANOVA (F = 1.2, p = 0.08), a PLS-DA model (CER = 0.32, Extended Data Fig. 5j ) identified several acylcarnitines as discriminant features between treated and untreated water ( Supplementary Table 1 ). Notably, we found an opposite pattern than we observed for infant feces, with acylcarnitines being depleted in human milk in the untreated water group (exemplified for acetylcarnitine (p < 0.01) and propionylcarnitine (p < 0.01) in Extended Data Fig. 5k ). These observations raise an intriguing hypothesis: that water treatment methods may have a direct impact on metabolic profiles. While this warrants further investigation, it is possible that water treatment serves as a proxy for broader socioeconomic differences, particularly given similar patterns observed with the asset index used as a socioeconomic proxy. Water treatment was weakly associated with overall gut microbiota composition, but it was linked to changes in specific genera, particularly in early life ( Fig 3k,l ). Infants in the untreated water group had enriched amounts of genus Flavonifractor (some members formerly Clostridium) , known for its ability to metabolize flavonoid polyphenols 53 , and a current Clostridium genus, which is known to play a role in the metabolism of carnitines and its related metabolites 54 . Consistent with this, a search of the spectra from discriminant acylcarnitines using microbeMASST 55 , identified matches to bacteria including Clostridium (exemplified for butyrylcarnitine in Extended Data Fig. 5l ). The elevated levels of carnitines in infant feces in the untreated water group may therefore be attributed to a microbiome enriched in species able to produce them, such as Clostridium species. However, it is also possible that higher availability of acylcarnitines in untreated water promotes the growth of carnitine-metabolizing bacteria. We also found that infants in the untreated water group had significantly lower phylogenetic alpha diversity in the first 30 days of life (2MWW-HB, p = 0.0072; Extended Data Fig. 5m ), suggesting a potential negative impact on the establishment of a diverse microbiome, which may have implications for long-term health given that lower microbial diversity has been associated with various diseases 56 , 57 . Maternal Secretor Status Influences Infant Fecal Metabolic Programming We also investigated the influence of maternal secretor status on both the human milk and the infant fecal metabolome. As expected and confirmed by targeted quantification using high-performance liquid chromatography (HPLC) with fluorescent derivatization, there was a clear difference in HMO milk profiles between secretors and non-secretors ( Extended Data Fig. 6a-b ), consistent with previous findings 58 – 60 . Surprisingly, unsupervised dimensionality reduction of the milk metabolome did not reach statistical significance for maternal secretor status, but a trend was observed (PERMANOVA: R 2 = 0.01, F-statistic = 1.2, p = 0.067). Nevertheless, by plotting the natural log ratio of discriminant features associated with secretor vs. non-secretor identified by PLS-DA (CER = 0.36; Extended Data Fig. 6c ) we observed a consistent separation over time (LME: β = 1.0, p = 5.1×10 -13 ; Fig. 4a ). The difference was largely driven by features putatively classified as saccharides by SIRIUS 61 , 62 . By molecular networking and manual inspection of MS/MS spectra, we identified several putative HMOs, including 2’FL ( m/z 471.1709) and difucosyllactose (DFLac) ( m/z 657.2211), another secretor-associated HMO 63 , 64 . As expected, these were detected almost exclusively in milk samples from secretors ( Fig. 4b ). While saccharides such as 2’FL cannot be unambiguously differentiated from their structural isomers by MS/MS alone, their exclusive presence in secretors supports their annotation as α1-2-fucosylated HMOs. Additionally, the observed patterns were consistent with the targeted quantification of 2’FL and DFLac ( Fig. 4b ), and in secretor milk, the targeted data correlated with the untargeted data for these compounds (Pearson correlation: 2’FL: r = 0.65, p = 1.8×10 -5 and DFLac: r = 0.84, p = 4.2×10 -10 ; Extended Data Fig. 6d ). Download figure Open in new tab Fig. 4: Maternal secretor status shapes HMO profiles in human milk and infant feces. a, Longitudinal natural log ratios of the summed peak areas of the top 70 discriminating features for maternal secretor status in human milk. Statistical significant difference between groups was assessed with LME. Scatter plot smoothing (‘LOESS’) added for visualization. b, Boxplot of metabolic feature ( m/z 471.1709) predicted to be 2’FL, targeted quantification of 2’FL, metabolic feature ( m/z 657.2211) predicted to be DFLac, and targeted quantification of DFLac, respectively, grouped by secretor status (Wilcoxon rank-sum test, with BH adjusted p values). For subjects with multiple samples within a stage, only the sample with the highest 2′FL concentration was retained. c, PCA of infant fecal samples showing separation by maternal secretor status (PERMANOVA: p = 0.006). d, Longitudinal natural log ratios of the summed peak areas of the top 100 discriminating features for maternal secretor status in infant feces, with the significance of the difference assessed by LME and smoothing for visualization. e, Piechart showing the distribution of NPC predicted chemical classes or pathways (probability > 0.7) for the discriminant infant fecal features by maternal secretor status. Only features with VIP > 2.5 were included. f, Cross-biofluid molecular network of amino sugars, aminoglycosides, and saccharides, as predicted by CANOPUS, colored by biofluid and secretor status. Node size represents the mean VIP score of significant features across the two PLS-DA models for secretor status in human milk and infant feces, with larger nodes indicating higher VIP score. Selected features enriched in both maternal milk and infant feces are shown in boxplots. g, Boxplot of the natural log-transformed peak areas of a candidate trihydroxy bile acid (Δ 163.0841) grouped by secretor status (Wilcoxon rank-sum test with BH adjusted p values). h, Boxplot of the natural log-transformed peak areas of putative 3α,7α-(OH) 2 -12-oxo based on secretor status at select timepoints (Wilcoxon rank-sum test with BH adjusted p values). All boxplots show the first (lower), median, and third (upper) quartiles, with whiskers 1.5 times the interquartile range. For boxplots in f, g, and h , repeated measures were avoided by retaining only the highest value per individual within each time window. For a and d, β and p values were derived from LME models with subject as random effect. The difference in human milk composition between secretors and non-secretors drove a corresponding association with the infant fecal metabolome (PERMANOVA: F-statistic = 1.5, p = 0.006; Fig. 4c ). The natural log ratio of discriminant features associated with secretor status identified by PLS-DA (CER = 0.34; Extended Data Fig. 6e ), revealed that infants of secretor mothers had distinct fecal metabolic profiles from those of non-secretor mothers in the first months of life (LME: β = 0.099, p = 1.6×10 -5 ; Fig. 4d ). The metabolic difference diminished around 3-4 months of life, likely reflecting reduced breastfeeding, as less than 45% of infants were still exclusively breastfed at that time. Key discriminant features included not only saccharides (11.4%), but also steroids (17.1%), sphingolipids (11.4%), and aminosugars/aminoglycosides (5.7%) (putatively classified by CANOPUS; Fig. 4e ), suggesting a broad effect of human milk composition on gut metabolic function. For example, dipeptides and microbial bile amidates, including glutamate-trihydroxy and glutamine-trihydroxy, were enriched in the feces of infants born to secretor mothers ( Supplementary Table 1 ). N-acetyl-D-glucosamine (GlcNAc), an amino sugar known to be an important energy source for gut bacteria 65 , were enriched in infants of non-secretors, supporting previous findings of lower GlcNAc in infant feces from secretor mothers 66 . Furthermore, acylcarnitines, histamine-trihydroxy bile acid, several short-chain N -acyl lipids, and polyamine derivatives (e.g. spectral matches to acetylputrescine-C2:0, cadaverine-C3:0, acetyl-palmitoyl-putrescine, and diacetylspermidine) were also enriched in non-secretors, whereas N -acyl lipids enriched in secretors were predominantly longer-chain compounds (C16-C18) ( Supplementary Table 1 ). Although the source tracking analysis showed that direct microbial seeding from maternal milk is limited, these findings highlight the indirect effect of HMOs in shaping gut metabolic programming 67 . A cross-biofluid molecular network demonstrated transfer of specific HMOs from human milk to infant feces, showing that HMOs abundant in secretor milk, including putative 2’FL and DFLac, were also enriched in the feces of respective infants ( Fig. 4f ). This suggests that maternal secretor status directly shapes the infant fecal metabolome through different composition of the milk itself, rather than simply differences in breastfeeding patterns. Interestingly, secretor status also influenced bile acid metabolism. Notably, a candidate trihydroxy bile acid (Δ 163.0841) was consistently elevated in infants of secretor mothers throughout the study period (Wilcoxon rank-sum test: all p < 0.01), except during the first postnatal week ( Fig. 4g ). We identified it as a bile acid given the two diagnostic ions, m/z 319.24 and 337.25, in the MS/MS spectrum 30 , with delta mass molecular formula C 6 H 13 NO 4 , as predicted by SIRIUS. We do not yet know the structure of the bile acid steroid core or the modification as it is a spectral match to the candidate bile acid library 30 and therefore it is referred to as a candidate bile acid. Conversely, the microbial-derived 3α,7α-(OH) 2 -12-oxo bile acid, characterized by a ketone at the C12-position and formed through oxidation by bacterial 12α-hydroxysteroid dehydrogenases 68 , exhibited a transient difference, with higher levels in infant feces of non-secretors between days 14 and 28 postpartum (Wilcoxon rank-sum test, p = 0.007; Fig. 4h ). Secretor status did not influence the infant fecal microbiome alpha diversity (Faith’s PD, LME: β = 8.1, p = 0.13) nor the fecal log ratio from Fig. 2c (LME: β = 0.001, p = 0.94; Extended Data Fig. 6f ). Identification of Microbial Drivers of Early-life Metabolic Programming through Multi-omics Analysis Finally, to identify candidate bacteria associated with the production of metabolites of interest, we performed a multi-omics analysis using joint-RPCA 36 . The joint analysis was restricted to infant fecal samples collected between days 10-30 after birth, roughly aligning with the earliest time window showing group separation by the variables investigated. We focused on microbial-modified bile acids, N -acyl lipids, acylcarnitines, and other key differential metabolites identified in earlier analyses. Microbiome features were selected based on the overlap between three analyses (ANCOM-BC2, joint-RPCA, and TEMPTED). The joint-RPCA found delivery mode as a significant factor shaping combined multi-omics analysis (PERMANOVA: delivery mode, p = 0.01, F-statistic = 3.98; Extended Data Fig. 7a ). Further, there were strong positive or negative correlation patterns with key features of interest ( Fig. 5a ; complete heatmap in Extended Data Fig. 7b ). Notably, several compounds associated with fatty acid metabolism, including microbial bile amidates, 2-aminovaleric acid-C2:0, valine-C2:0, and several acylcarnitines negatively co-occurred with OGUs belonging to Escherichia and Bifidobacterium genera. This may indicate that these molecules can be used by these genera for growth or other metabolic functions. Consistent with other reports 16 , 17 , we observed a positive correlation between Bifidobacterium and HMOs. Preliminary evidence suggests that Escherichia and Bifidobacterium can act cooperatively to digest HMOs 69 . Furthermore, since Bifidobacterium spp. are critical members of the early-life gut microbiome 70 , 71 , and are involved in HMOs utilization, it was interesting to note that they also displayed strong positive co-occurrences with biliverdin, several acylcarnitines, 3α,7α-(OH) 2 -12-oxo bile acid, and at least one sulfated trihydroxy bile acid. This suggested that Bifidobacterium spp. may produce or co-metabolize these compounds. To assess this further, we searched the MS/MS spectra of these metabolites with microbeMASST 55 . Overall, we obtained a match for 68% of the searched spectra, suggesting that these bacteria could actually produce the molecules of interest. Importantly, 53% of them had matches to monocultures from the same genera to which they were associated in the heatmap ( Fig. 5a ). For example, acetylcarnitine (C2:0) was frequently observed in cultures of Bifidobacterium ( Extended Data Fig. 7c ). Download figure Open in new tab Fig. 5: Multi-omics integration identifies microbial mediators of metabolic trajectories. a, Heatmap of correlating metabolites and microbial species from infant fecal samples collected at day ∼30 identified by joint-RPCA. Metabolic features shown were selected based on previous analyses. Asterisk indicate MS/MS spectral match against microbeMASST, with (*) indicating a general match and (**) indicating a match to a monoculture of the same genus as taxa shown in the heatmap. Microbiome features shown were selected based on the overlap between 3 previous analyses, including needing to belong to the differentially abundant genera found in the ANCOM-BC2 analysis from Figure 2b , the top and bottom 20 most differentially ranked features for joint-RPCA Axis 1, and the top and bottom 10% differentially ranked features from TEMPTED Axis 1 (overlapping features=23). DISCUSSION This study, including over 1,600 metagenomics samples and more than 500 untargeted metabolomics profiles across multiple body sites, offers unique and comprehensive insights into infant metabolic programming during early life in a cohort from Bangladesh. Through frequent sampling over the first six months of life, we profiled the metabolome and microbiome of 55 mother-infant dyads. Leveraging recently developed MS/MS spectral libraries 27 , 30 – 32 , we describe the fecal developmental trajectories of key signalling molecules, including bile acids, N -acyl lipids, and acylcarnitines. Through multi-omics integration we also reveal specific microbial-metabolite relationships, emphasizing the role of the gut microbiome as an important mediator of early developmental metabolic trajectories. We show that the functional capability of the gut microbiome matured rapidly, demonstrated by an early rise in microbially derived bile acids and short-chain N -acyl lipids in the first weeks after birth. This reflects the developing microbial proficiency in performing modifications of these compounds. The decline in host-derived taurine and glycine conjugates alongside the rise in microbial bile acids aligns with previous findings 33 , 72 , suggesting a shift toward a microbiome capable of more complex bile acid transformations, such as deconjugation and amidation 30 . This was further supported by increased levels of the bsh genes with infant age, reflecting colonization of bsh-expressing bacteria as the infant grows 33 . Understanding the developmental trajectories of these compounds is crucial as bile acids not only interact with nuclear receptors such as farnesoid X receptor (FXR) 33 , 73 and pregnane X receptor (PXR) 31 , 74 , but also modulate immune processes, including T-cell differentiation and dendritic cell signaling 31 , 33 , 74 , thereby shaping and programming infant immune development. Further, we demonstrate that the trajectories of these metabolites are sensitive to factors like delivery mode and maternal secretor status. Despite only minor differences in infant gut microbiota composition, infants born via C-section had unique metabolic profiles, with transiently higher levels of bile amidates and N -acyl lipids, as well as short-chain acylcarnitines, during the first month of life. Beyond their critical role in energy metabolism and mitochondrial homeostasis 52 , 75 , acylcarnitines have been associated with rapid infant growth 70 , and they are also involved in immune homeostasis 52 by modulating inflammatory signalling pathways, such as NF-κB and JNK 76 , 77 . Second, we found that maternal secretor status also influenced the bile acid pool, with levels of a candidate trihydroxy bile acid (Δ 163.0841) and 3α,7α-(OH) 2 -12-oxo bile acid differing between infants of secretor and non-secretor mothers. This suggests that the maternal HMO profiles, through the infant gut microbiome modulation 17 , 67 , 78 , may shape early-life bile acid metabolism. Source tracking analysis revealed that, while the maternal fecal microbiome is the main contributor to the infant fecal microbiome, environmental sources also play a major role in shaping its composition. Environmental exposures such as drinking water quality can differ substantially between LMIC and high-income settings based, for example, on contamination with diarrheagenic 79 or antibiotic-resistant pathogens 80 . Exogenous elements, including xenobiotic trace elements in water, have been linked to changes in the diversity and composition of the gut microbiome in both adults 50 , 51 and infants 81 . Here, we show that differences in the infant fecal metabolome and microbiome are associated with reported household-level drinking water treatment practices, specifically in relation to samples collected during the first month of life. Given that infants usually do not directly consume water at this age, this variable should be interpreted as a proxy for household hygiene and sanitation related practices. As it is also likely associated with socioeconomic status and other unmeasured factors, this finding may reflect a broader constellation of hygiene practices, infrastructure, and living conditions, rather than water treatment per se . Notably, infants in households using untreated water presented higher fecal levels of acylcarnitines, despite being depleted in the milk of their respective mothers. Accumulation of long-chain carnitines in the feces has been previously associated with intestinal inflammation in the pediatric context and growth promotion of Enterobacteriaceae species and other pathobionts 82 . Supporting this, we found Clostridium species, also known to metabolize acylcarnitines 54 , enriched in infants from households that did not report using methods to treat water. Elevated levels of acylcarnitines have been linked with negative health outcomes 83 – 85 , but long-term health implications remain to be investigated. Another exposure that can differ substantially from high-income settings is the widespread use of antibiotics in LMICs 21 – 23 . This creates a different baseline for microbial development than in a Western setting, where treatment with antibiotics is less frequent 86 , potentially explaining the observed differences in alpha diversity for infant gut microbiota 41 . In conclusion, this longitudinal multi-omics study emphasizes the influence of delivery mode, maternal secretor status, and environmental factors like water treatment in shaping the early infant microbial and metabolic profiles in a LMIC setting. This study has a few limitations, including a relatively small sample size for the dyads (n=55), COVID-19 pandemic may have impacted household habits during the study, reliance on self-reported data for variables like water treatment, widespread antibiotic usage, the lack of inclusion of seasonality in the analysis, and metabolite annotations based on MS/MS spectral matches with the GNPS spectral libraries (level 2-3). Additionally, the water source was not considered, but most households in Dhaka used piped city water or deep tube wells 87 . Despite these limitations, we show that the infant gut microbiome, a dynamic ecosystem, requires a multi-omics approach to better understand the rapidly changing metabolic mechanisms of early life that can ultimately impact the future health and development of individuals. Follow-up studies using long-read sequencing, which provides better strain-level resolution 88 , in larger cohorts will be essential for determining the extent and nature of the observed responses to both strain transmission as well as water quality. METHODS Ethics statement Ethics approval for the SEPSiS (Synbiotics for the Early Prevention of Severe Infections in Infants) Observational Cohort Study, from which the biological specimens and data used in this study were derived, was obtained from the Research Ethics Board of The Hospital for Sick Children (No.: 1000063899) and the International Center for Diarrhoeal Disease Research, Bangladesh (icddr,b; PR-19045). In addition, this project has been reviewed by the Director of the UCSD HRPP, IRB Chair, or IRB Chair’s designee and is exempt from Institutional Review Board requirements under categories 45 CFR 46.104(d), category 4: Secondary research for which consent is not required: Secondary research uses of identifiable private information or identifiable biospecimens; ii) Information, which may include information about biospecimens, is recorded by the investigator in such a manner that the identity of the human subjects cannot readily be ascertained directly or through identifiers linked to the subjects, the investigator does not contact the subjects, and the investigator will not re-identify subjects. Study design and recruitment Mother-infant pairs were screened at two government health care facilities offering delivery services in Dhaka: Maternal and Child Health Training Institute (MCHTI) and Mohammadpur Fertility Services and Training Centre (MFSTC). Enrollment began in May 2021 and the last study visit was completed in April 2022. Eligible participants were offered enrollment into ‘Schedule A’ (n=75) of the SEPSiS Observational Study, which was designed for metagenomics and metabolomics analyses ( Extended Data Fig. 1a ). Infants were eligible for inclusion if: informed consent was received by parent or guardian, they were between 0-1 days old at screening, delivered at one of the study hospitals, orally feeding at the time of screening and planned to reside within the study catchment area until 60 days of age. Exclusion criteria were: birth weight <1500 g, major surgery required within the first week of life, and/or major congenital anomaly of the gastrointestinal tract, maternal human immunodeficiency virus (HIV) infection and/or history of mother ever receiving antiretroviral drug(s) for presumed HIV infection or administration/prescription of parenteral antibiotics or any prenatal or postpartum use of non-dietary probiotic supplement, infants that required mechanical ventilation and/or cardiac support, infants actively enrolled in clinical trials that included the use of probiotics or prebiotics, infants residing in the same household as another infant previously enrolled in the study, or any study within the research platform, who is currently <60 days of age. Field activities were overseen by the International Centre for Diarrheal Disease Research, Bangladesh (icddr,b). Participants were followed from enrollment to 6 months of age. A total of 55 mother-infant pairs were selected from the 75 infants enrolled in the SEPSiS Schedule A arm and included in this study. In addition to the general inclusion and exclusion criteria described in the study design section, infants were specifically selected to contribute samples to this study if the following additional criteria were met: if day 0 or 1, 60, 90 and 180 stool samples were collected and available, at least 11 routine clinical assessments were completed, 7 or more stool samples were collected, a minimum set of data collection modules relevant for the study analyses were completed (e.g., feeding practice, birthweight, gestational age information). Enrollment was stratified by delivery mode to ensure a similar proportion of infants delivered vaginally and via C-section. Sample collection Samples were collected longitudinally from various body sites by trained study personnel during routine visits in the hospitals until discharge or in the participant’s home after discharge. Eleven fecal samples were scheduled for collection from mothers and infants on days 0 or 1, 3 or 4, 6 or 7, 10, 14, 21, 28, 35, 60, 90, and 180, with allowances for rescheduling up until the next scheduled visit. Infant oral (tongue) and skin swabs (elbow) were scheduled at 8 different timepoints (day 0 or 1, 6 or 7, 14, 21, 35, 60, 90, 180), with allowances for rescheduling up until the next scheduled visit. Mothers provided a vaginal swab at enrollment (day 0/1). Maternal milk was collected at enrollment (0/1), day 60, and 2 other randomly determined timepoints up to day 60 of the infant’s age. Infant blood samples were collected at day 1 of age only. Infant stool samples were collected by trained study personnel, aliquoted and stored in the vapour phase of liquid nitrogen within 20 minutes of defecation. Maternal fecal samples were collected and placed in a cold box within 20 minutes of defecation and then homogenized, aliquoted and frozen by a study worker within 6 hours of defecation. Vaginal swabs were self-collected by mothers at each study hospital after delivery following the instructions provided by study nurses and physicians. Mothers hand-expressed human milk in sterile containers which were aliquoted and stored on ice. Infant oral (from the tongue dorsum) and skin swabs (from the antecubital fossa, on the inner left or right elbow) were collected aseptically by trained study workers and placed in sterile containers. One routine venous blood sample was collected from infants by a trained phlebotomist, study nurse or medical officer. All specimens were transported to the International Centre for Diarrheal Disease Research, Bangladesh (icddr,b) and stored in -70°C until they were transported on dry ice from Dhaka to University of California San Diego (UCSD) for metagenomic, metabolomics, and HMO laboratory analysis. Main study variables and definitions To assess contribution of various maternal and infant characteristics to the variation in the infant fecal metabolome we performed a permutational multivariate analysis of variance (PERMANOVA) 89 . Infant age was the primary variable of interest to describe developmental trajectories of key signalling molecules. In addition to infant age, the PERMANOVA model included variables selected for their known influence on the early life gut microbiome and metabolome: delivery mode (categorical) 6 , 8 , maternal secretor status (categorical) 17 , 67 , gestational age (continuous) 90 , feeding patterns (categorical) 91 , 92 , infant sex (categorical) 93 , birthweight (continuous) 93 , antibiotic usage by mother and by infant (categorical) 7 , 20 , socioeconomic status (categorical) 94 , 95 , and household water treatment (categorical) 50 , 51 . Based on the PERMANOVA output, three significant covariates were prioritized for in depth-analysis: delivery mode, maternal secretor status, and household water treatment. - Delivery mode: categorized as vaginal or C-section based on clinical records. - Maternal secretor status: categorized as secretor or non-secretor based on presence or near-absence (<100 nmol/mL) of HMO 2’-fucosyllactose. - Household drinking water treatment : this binary variable was defined by self-reports collected at baseline. Households were categorized as ‘treated’ if they reported purifying their drinking water (either using boiling and/or filtering) or ‘untreated’ if they reported no treatment of their drinking water. Additionally, we briefly assessed the impact of socioeconomic status on the infant fecal metabolome using an asset index. The asset index and quintiles were generated using PCA for all participants enrolled in the SEPSiS Observational Study (n=1,939) and an interventional trial (n=519) at the same study sites with the same eligibility criteria 25 . The score was a summary measure of household wealth based on ownership of the following items: electricity, almirah, fan, table, chair, radio, television, fridge, pump, freezer, mobile, watch, computer, phone, animals, autobike, bicycle, rickshaw and a vehicle, where higher scores correspond to greater wealth. The resulting distribution was split into quintiles (Q1-Q5), leading to an uneven distribution of subjects in each quintile for our study cohort. For statistical analysis, the asset index was converted into a binary variable by grouping the lowest quintiles (Q1-Q2) and the highest quintiles (Q3-Q5). For maternal milk, infant age and maternal secretor status were the primary variables of interest. We also assessed metabolic changes across lactation stages, by grouping milk samples into colostrum ( 15 days). Untargeted metabolomics sample preparation Metabolomics analysis was conducted on 380 infant fecal samples, 34 infant plasma samples, and 119 mother’s milk samples. Stool and plasma were extracted using 95% ethanol (EtOH) via the Matrix tube method as previously described 96 , 97 . Briefly, samples were extracted in 400 µL of 95% EtOH, shaked at 1,200 rpm for 2 min in a SpexMiniG plate shaker (SPEX SamplePrep part #1600, NJ, USA), and centrifuged for 5 min at 2,700 g. Supernatant (200 µL) was then collected and stored at -80°C for later acquisition. Prior to instrumental analysis, plates were redissolved in 200 µL 50:50 methanol (MeOH):H20. Milk samples were extracted in 75:25 MeOH:methyl tert-butyl ether (MTBE) as previously described 98 . Briefly, 400 µL of the MeOH/MTBE solution was added to 50 µL of milk, sonicated for 5 min, followed by a 15 min centrifugation (4°C) at 1800 rpm. Supernatant (300 µL) was immediately transferred to a new 96 well plate for instrumental analysis. Untargeted metabolomics data acquisition Fecal and plasma samples were randomized and run on the same batch, while mother’s milk samples were run on a separate batch. Samples (5 μL) were injected into an UltiMate 3000 liquid chromatography system (Thermo Scientific) coupled to a QExactive Orbitrap (Thermo Scientific) mass spectrometer. Chromatography was conducted using a Kinetex C18 column (Phenomenex). A representative linear gradient of mobile phase A (water + 0.1% formic acid) and phase B (acetonitrile + 0.1% formic acid) consisted of 0-1 min isocratic at 5% B, 1-7 min to 98% B, 7-7.5 min isocratic at 98% B, 7.5-8 min to 5% B, and 8-10 min at 5% B with a flow rate of 0.5 mL/min. All solvents used were LC-MS grade. Data was acquired in data-dependent acquisition (DDA) mode using tandem mass spectrometry (MS/MS) in positive electrospray ionization (ESI+). Briefly, full scan MS spectra was acquired at 35,000 resolution with an automatic gain control (AGC) target of 5e5, maximum ion injection time of 100 ms, and a scan range of 100-1500 m/z . MS/MS spectra were collected using the same resolution and AGC target, and fragmented the top 5 most abundant ions per cycle with a 3.0 m/z isolation window, stepped normalized collision energies of 20, 30, and 40%, and a dynamic exclusion window of 10 s. Untargeted metabolomics data processing Acquired .raw files were converted into .mzML format using ProteoWizard MSConvert 99 and deposited in GNPS/MassIVE under the accession code MSV000096943 . Feature detection and extraction were performed using MZmine 4.2 100 . Briefly, data were imported with MS1 and MS2 detector noise factor set to 3 and 2 respectively. The chromatogram builder parameters were set at 5 minimum consecutive scans, 1e5 minimum absolute height, and 10 ppm for m/z tolerance. Smoothing and local minimum feature resolver were applied, with the following parameters for the latter: chromatographic threshold 85%, minimum search range retention time 0.2 min, and minimum ratio of peak top/edge 2. Isotope filter and finder were then applied before feature alignment via join aligner (3 weight for m/z, 0.3 retention time tolerance). Peak finder was performed after removing features that were not detected in at least 3 samples. Further, metaCorrelate and ion identity molecular networking were run. List of MS/MS spectra and associated feature table was then exported and used to generate a feature based molecular network (FBMN) 101 on GNPS2 102 . The FBMN job is publicly available at https://gnps2.org/status?task=d3c90a3743cf435291b635404821ab53 . Briefly, fragment tolerances for both parent and fragment ions were set at 0.02 and networking and annotation parameters were both set to 5 minimum matching peaks and cosine similarity > 0.7. The network.graphml file from the FBMN was imported and manipulated in Cytoscape 103 (version 3.10.3). MS/MS spectra of unknown features were also run in SIRIUS 61 (version 6.1.1) and CANOPUS 62 was used to predict features’ chemical pathway, superclass or class. Features of interest were searched using microbeMASST 55 ( https://masst.gnps2.org/microbemasst/ ) (minimum 4 matching peaks and cosine similarity > 0.7; output from 07/15/2025), a taxonomically informed mass spectrometry search tool within GNPS where MS/MS spectra can be linked to their respective microbial producers by querying against a curated reference database of >60,000 microbial monocultures. Untargeted metabolomics data analysis The feature and annotation table were imported in R 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria) for downstream data analysis. Sample’s total extracted peak areas were investigated to look into poor data acquisition due to sample collection, degradation, or instrument variation. Features with a retention time (RT) 8 min were excluded. Blank subtraction was performed by removing features detected in Blank samples meaning peak areas were not at least 5 times that observed in the samples. The package ‘homologueDiscoverer v 0.0.0.9000’ was used to remove polymers detected in the LC-MS/MS run 104 . Three fecal samples were excluded from the metabolomics analysis due to missing metadata, unusually high intensity in the ion total chromatogram, and clustering with the plasma samples. Features with near zero variance were removed using the package ‘caret v 6.0’. Features annotated as bile acid candidates using the propagated bile acid library 30 underwent post-validation filtering based on the presence of diagnostic MS/MS fragment ions using massQL 105 ( https://massqlpostmn.gnps2.org/ ). Features were robust center log ratio (RCLR) transformed using ‘vegan v 2.6’ before applying dimensionality reduction techniques. Principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA) models were built using ‘mixOmics v 6.22’. PERMANOVA was used to test centroid separation. The performances of the PLS-DA models were investigated via a 5-folds cross-validation. VIP scores > 1 were considered significant for group separation. The natural log ratios were generated by summing the peak areas of features enriched in one condition (numerator) relative to features enriched in the other condition (denominator). For compounds with multiple annotations, e.g. due to isomers producing similar MS/MS, peak areas were summed when plotting their trajectories over time (e.g. this was the case for some N -acyl lipids). The ‘lmerTest v 3.1’ package was used to generate the linear mixed effect models, where subject id was used as random effect and infant age as a fixed effect. For panels showing trajectories with age, smoothed trend lines using ‘LOESS’ (non-parametric local regression smoothing method) were added for visualization purposes, and were not derived from LME. The Wilcoxon rank-sum test was used to determine statistically significant group differences followed by Benjamini-Hochberg correction with p value < 0.05 considered statistically significant. To avoid repeated measures in comparisons across time windows (e.g. days 0-7, 14-28, and 30-45), only the sample with the highest value (peak area or log ratio) per subject within each window was retained. For the subanalysis investigating discriminant features between days 10-30 for delivery mode, one fecal sample per infant was selected, specifically the sample with the highest log ratio within this interval. Features detected in > 20% of the samples were included. Log₂ fold changes were calculated between groups and features with an FDR-adjusted p value < 0.10 were considered relevant. Upset plots were generated using the package ‘UpSetR v 1.4’ 106 . Only human milk from Secretors (functional FUT2 gene) was included in the correlation analysis between the targeted quantification of HMOs using HPLC after fluorescent derivatization and the untargeted LC-MS/MS data. Zero values, only present in the untargeted data, were also excluded from the analysis (46% of samples for 2’FL and 54% of samples for DFLac). Sample processing for metagenomic sequencing All samples were transferred into 1 mL Matrix Tubes (ThermoFisher, Waltham, MA, USA) for gDNA extraction. Samples were extracted for metagenomic sequencing using reagents from the MagMAX Microbiome Ultra Nucleic Acid Isolation Kit (ThermoFisher Scientific, Waltham, MA, USA) following a previously described high-throughput protocol 107 , minimizing reaction volumes, eliminating the need for the sample-to-bead plate transfer step, and reducing potential cross-contamination from vortexing during bead beating. Extracted gDNA was quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Waltham, MA, USA) and normalized to 5 ng in 3.5 µL sterile water for library preparation, which was performed using a miniaturized adaptation of the KAPA HyperPlus Library Kit (Roche, Basel, Switzerland) 108 . The library was quantified via PicoGreen Assay (Invitrogen), and all samples were equal volume pooled, polymerase chain reaction (PCR) cleaned (Qiagen), and size selected from 300 to 700 bp using a Pippin HT (Sage Sciences). QC was run on an Agilent 4000 Tapestation (Agilent, Santa Clara, CA, USA) to confirm expected library sizes after PCR cleanup and size selection. The equal volume pool was sequenced on an iSeq100 (Illumina, San Diego, CA, USA). Utilizing the sample concentration and read counts per sample obtained from the iSeq 100 run, a normalized pooling value was calculated for each sample to optimize pooling efficiency to obtain more even read counts per sample during NovaSeq sequencing 109 . After re-pooling the library with the iSeq normalized pool values, samples were PCR cleaned, size selected (300–700 bp), and QC was performed using an Agilent tapestation. Sample bioinformatic analysis The PCR-cleaned, size-selected, iSeq-normalized pool was sequenced on a NovaSeq 6000 (Illumina, San Diego, CA, USA) at the Institute for Genomic Medicine at the University of California, San Diego with an S4 flow cell and 2 × 150 bp chemistry. Raw sequence reads (BCL files) were demultiplexed to per sample FASTQ and quality filtered 110 . Adapter trimming was performed by fastp 111 , and human reads were filtered out to ensure compliance with database regulations. This was performed via the conservative host filtration pipeline described previously 112 . Briefly, reads were first filtered using minimap2 113 via alignment to two reference genomes: GRCh38.p14 plus PhiX, and CHM13v2.0. Additionally, a pangenome index was created using the 94 currently available pangenomes from the Human Pangenome Reference Consortium 114 using Movi 115 . Unmapped reads following alignment were queried against this index, and reads with high-propensity matches to one or more human pangenome references were discarded. The resulting FASTQ files were uploaded into Qiita 116 ( study ID #14834 ) and processed using the Woltka pipeline 117 (version 0.1.7) using SHOGUN 118 parameters. In brief, direct genome alignments were made against the “Web of Life” database (release 2) 119 , which contains 15,953 microbial genomes of bacteria and archaea. Paired-end sequence alignment is performed using the bowtie2 (v2.5.4) aligner 120 (using --very-sensitive -k 16 --np 1 --mp "1,1" -- rdg "0,1" --rfg "0,1" --score-min "L,0,-0.05" --no-head --no-unal --no-exact-upfront --no-1mm- upfront) and by mapping sequencing data to microbial reference genomes 118 . Microbial genome IDs are considered operational genomic units (OGUs) 117 . The OGU frequencies were then summed after the entire alignment was processed and rounded to the nearest even integer, thereby making the sum of OGU frequencies per sample nearly equal (considering rounding) to the number of aligned sequences in the data set. The resultant count matrix is saved and formatted as a BIOM format table 121 , which has microbial genome IDs (rows) by samples (columns) that contains read counts as values. The BIOM file was converted to a QIIME2 122 (v2024.5) artifact. We applied a novel coverage dispersion filter called micov which eliminated OGU with less than 25% reference genome coverage ( https://github.com/biocore/micov ) to reduce risk of false positives from short read multimapping. A 1% prevalence filter was also applied. Contamination in the resultant table was removed using SCRuB (Source-tracking for Contamination Removal in microBiomes) 123 , a probabilistic decontamination method that incorporates samples and controls to remove contamination. Due to the large sample size of this study SCRuB was run separately for each different sample plate to account for unique well-to-well contamination. The resulting table contained 1,568,600,606 reads for 1,609 samples, with an average of 901,495 reads per sample. Standard alpha diversity metrics (Faith’s PD 124 , Shannon index 125 ), as well as standard beta diversity metric, unweighted UniFrac 126 , and phylogenetic robust center-log-ratio (rclr) metric, phylo-RPCA 36 , were calculated. For unweighted UniFrac, dimensionality was reduced using PCoA and visualized using EMPeror 127 . For phylo-RPCA, dimensionality was reduced using PCA and visualized using EMPeror. For unweighted UniFrac, the feature table was rarified according to a read count value that would retain 98.5% of all samples from all body sites (including 90% of vaginal samples), 644 reads. Alpha diversity rarefaction depth for low biomass samples (vaginal, human milk) is 644, high biomass samples is 100,000 (fecal, oral, skin). For rclr-based diversity metrics (phylo-RPCA), an unrarefied table (sample minimum read count for inclusion was 644 reads for cross body site comparison in Extended Data Fig. 4. When binned timepoints are used, samples are grouped into Day 0, 7, 14, 30, 60, 90, 180 to which they are closest. TEMPoral TEnsor Decomposition (TEMPTED) 128 is a beta diversity dimensionality reduction method that accounts for repeated sampling and uses time as a continuous variable, plus it can handle varying temporal sampling present in the dataset. Beta diversity differences were determined by PERMANOVA. Feature tables were collapsed to the genera level and then differential abundance was performed using analysis of compositions of microbiomes with bias correction (ANCOM-BC2) 129 . ANCOM-BC2 utilized the binary age category (greater than 1 month vs. less than 1 month of age) in the differential abundance modeling formula. The differentially abundant genera were used to create a natural log ratio of the sum of all read counts from genera enriched after one month (numerator) to all genera depleted (denominator). Qurro was used to visualize ranked features 130 . ConvexHull (CH) volume and area are measures that capture the dispersion or spread of a set of points in a multidimensional space (scipy ConvexHull v0.12.0). CH volume has been used in the past to assess the stability of microbial communities over time 131 . Here, a convex hull volume or area is calculated over all samples of a given body site at a given time. We also employed several linear mixed-effects models (LME) using the statsmodels package in Python to assess whether dependent variables of interest (faith pd values, convex hull volumes, etc.) varies with fixed effects. Generally, infant age at time of sample (days postpartum) was used as a fixed effect and individual host subject identification was used to account for repeated measures. For example, most of the alpha diversity metrics’ LME formulas were “faith_pd ∼ host_age_infant + (1 | host_subject_id)”. For Fig 1h , the formula was “ln( bsh reads) ∼ ln(AA/Tau conjugates) + (1 | host_subject_id)”. For Extended Data Fig. 6f , the formula was “[faith_pd or log_ratio] ∼ host_age_infant * hmo_secretor + (1 | host_subject_id)”, the star indicates an interaction term and is the one reported for significance. For panels showing trajectories with age, lineplots using binned timepoints for smoothing were added for visualization purposes and were not derived from LME. When binned timepoints were used in boxplots, differences were assessed using two-sided Mann-Whitney-Wilcoxon, followed by Holm-Bonferroni correction to adjust for multiple comparisons (2MWW-HB). Fast expectation-maximization microbial source tracking (FEAST) was used to estimate the contribution of potential source environments to a given sample 38 . Infant fecal samples were used as sinks in this analysis. The early timepoint used was from the birth/0 day sample, or if unavailable, the first sample collected for each infant. The late timepoint used the 6 month/180 day sample, or if unavailable, the final sample collected for each infant. For each timepoint dataset, we examined the estimated contribution proportion of matched or unmatched maternal biospecimens as potential sources. Using different_sources_flag = 0, all mother vaginal, fecal, and milk samples across all timepoints are given as potential sources. Using different_sources_flag = 1, all paired-mother vaginal, fecal, and milk samples across all timepoints are given as potential sources. Bile salt hydrolase/transferase gene search Gene searches were specifically performed for bile acid modification enzymes (hsdh, bai, bs h). For each gene of interest, faa files of all matches were acquired from Uniprot 132 , pooled, and then amino acid sequences with at least 60% similar identity were clustered via USEARCH 133 into a unique database. Then, metagenomic reads from each sample were aligned and matched to the genes of interest against the respective USEARCH clustered databases using Diamond blastx 134 . Output data frames were further filtered for matches with >40 bp of alignment length, > 50% query alignment ratio, and e-value< 0.00001 before plotting. Multi-omics analysis Joint-RPCA is an unsupervised method used to agnostically determine co-occurrences between the microbiome and metabolome data 36 . Since the method is not optimized for longitudinal sampling, samples around day 20 (day 10-30) were chosen for analysis. In short, the raw features tables for each method after previously described filtration methods were robust center log ratio (RCLR) transformed. RCLR of the raw observed values were only computed on the non-zero entries and then averaged and optimized. This helps to account for the sparsity and compositionality. The shared samples of all input matrices were used to estimate a shared matrix. The estimated shared matrix and the matrix of shared eigenvalues across all input matrices were recalculated at each iteration to ensure consistency. Minimization was performed across iterations by gradient descent. Cross-validation of the reconstruction was performed in order to prevent overfitting of the joint factorization. Co-occurrences of features across the two input matrices were calculated from the final estimated matrices. HMO analysis HMO profiles were obtained from 130 samples from 55 mothers at enrollment (0/1), day 60 and 2 other timepoints up to day 60 of the infant’s age. HMOs were analyzed using HPLC after fluorescent derivatization allowing for quantification of the 19 most abundant HMOs as previously described 135 , 136 . HMO concentrations are reported in nmol/mL, and the sum of all HMOs in a sample was calculated as the sum of all HMOs detected in each sample. HMO-bound sialic acid and HMO-bound fucose were calculated as the sum of sialic acid and fucose moieties included in each HMO. Overall, 19 HMOs were identified and quantified: 2′-fucosyllactose (2′FL), 3-fucosyllactose (3FL), 3′-sialyllactose (3′SL), 6′-sialyllactose (6′SL), difucosyllactose (DFLac), difucosyllacto- N -hexaose (DFLNH), difucosyllacto- N -tetrose (DFLNT), disialyllacto- N -hexaose (DSLNH), disialyllacto- N -tetraose (DSLNT), fucodisialyllacto- N -hexaose (FDSLNH), fucosyllacto- N -hexaose (FLNH), lacto- N -fucopentaose (LNFP) I, LNFP II, LNFP III, lacto- N -hexaose (LNH), lacto- N -neotetraose (LNnT), lacto- N -tetrose (LNT), sialyl-lacto- N -tetraose b (LSTb), and sialyl- lacto- N -tetraose c (LSTc). Data availability Untargeted LC-MS/MS data generated for this study are publicly available at GNPS/MassIVE ( https://massive.ucsd.edu/ ) under the accession code MSV000096943 . The associated FBMN job is publicly available at GNPS2: https://gnps2.org/status?task=d3c90a3743cf435291b635404821ab53 . Metagenomics sequencing data will be available through EBI/ENA via the following accession codes: PRJEB83236 ERP166886 (on publication). Additional information and processing pipelines will be available in Qiita Study ID #14834 ( https://qiita.ucsd.edu/study/description/14834 ). Code Availability Metagenomics analyses located at https://github.com/knightlab-analyses/gates_sepsis_bangladesh . Metabolomics analyses located at https://github.com/kinekvitne/manuscript_gates_sepsis_obs Funding This project was primarily supported by the Gates Foundation, INV-015983. This work was additionally supported by NIDDK 1R01DK136117-01 granted to P.C.D., and the Maternal and Pediatric Precision In Therapeutics (MPRINT) program project P50HD106463 granted to S.M.T. L.P. is supported by the University of California San Diego Medical Scientist Training Program (NIH/NIGMS T32GM007198). L.B. is the UC San Diego Chair of Collaborative Human Milk Research, endowed by the Family Larsson-Rosenquist Foundation, Switzerland. Reproductive Scientist Development Program Grant (K12HD000849) supports L.A.B. S.T. was supported by P50HD106463 and T32HD087978. Disclosures P.C.D. is an advisor and holds equity in Cybele, Sirenas, and BileOmix, and he is a scientific co-founder, advisor, holds equity and/or receives income from Ometa, Enveda, and Arome with prior approval by UC San Diego. P.C.D. consulted for DSM Animal Health in 2023. R.K. is a scientific advisory board member, and consultant for BiomeSense, Inc., has equity and receives income. R.K. is a scientific advisory board member and has equity in GenCirq. R.K. is a consultant and scientific advisory board member for DayTwo, and receives income. R.K. has equity in and acts as a consultant for Cybele. R.K. is a co-founder of Biota, Inc., and has equity. R.K. is a co-founder and has equity and is a scientific advisory board member of Micronoma, and has equity. R.K. is a board member of Microbiota Vault, Inc. R.K. is a board member of N=1 IBS advisory board and receives income. R.K. is a Senior Visiting Fellow of HKUST Jockey Club Institute for Advanced Study. The terms of these arrangements have been reviewed and approved by the University of California San Diego in accordance with its conflict of interest policies. D.M. is a consultant for, and has equity in, BiomeSense, Inc. The terms of these arrangements have been reviewed and approved by the University of California, San Diego. L.B. is a co-inventor on patent applications related to the use of HMOs in preventing NEC and other inflammatory diseases. All other authors declare no conflicts of interest. Author contributions K.E.K. conducted untargeted metabolomics analysis, multi-omics analysis, generated figures, and wrote the manuscript. C.A. conducted metagenomics analysis, multi-omics analysis, generated figures, and wrote the manuscript. J.C.O was involved in metadata preparation, conducted metagenomics analyses, and wrote sections of the manuscript. D.P. conducted metagenomics analysis and edited the manuscript. S.Z., V.C.L., I.M., and A.P. performed untargeted metabolomics analysis. S.T. processed untargeted metabolomics samples. K.S., L.P.,and L.B. conducted HMO analysis. A.G. and D.M. processed metagenomics samples. L.A.B. provided feedback on analysis, wrote portions of manuscript and edited manuscript. S.M.T. provided feedback and secured funding. L.G.P and D.E.R. designed the SEPSiS project from which data were used for the present study; provided feedback on the manuscript. L.B., P.C.D. and R.K. designed the study, provided feedback, and secured funding. All authors reviewed and approved the manuscript. SUPPLEMENTARY TABLE TITLES AND LEGENDS Supplementary Table 1. Output from the PLS-DA models showing variable importance in projection (VIP) scores across variables for maternal milk and infant feces: lactation stage (maternal milk), maternal secretor status (maternal milk and infant feces), delivery mode (infant feces), and household drinking water treatment (maternal milk and infant feces). Supplementary Table 2. Output from multivariate PERMANOVA. Supplementary Table 3. Overlapping discriminant features for water treatment identified by PLS-DA. Acknowledgements The authors would like to acknowledge the work of Gail Ackermann, who assisted with project management at UC-San Diego. Funder Information Declared Bill & Melinda Gates Foundation , INV-015983 NIDDK , 1R01DK136117-01 Maternal and Pediatric Precision In Therapeutics (MPRINT) , P50HD106463 University of California San Diego Medical Scientist Training Program, NIH/NIGMS , T32GM007198 Reproductive Scientist Development Program Grant , K12HD000849 NIH , T32HD087978 REFERENCES 1. ↵ Gensollen , T. , Iyer , S. S. , Kasper , D. L. & Blumberg , R. S . How colonization by microbiota in early life shapes the immune system . Science 352 , 539 – 544 ( 2016 ). OpenUrl Abstract / FREE Full Text 2. ↵ Renz , H. , Brandtzaeg , P. & Hornef , M . The impact of perinatal immune development on mucosal homeostasis and chronic inflammation . Nat. Rev. Immunol . 12 , 9 – 23 ( 2011 ). OpenUrl CrossRef PubMed 3. ↵ Dominguez-Bello , M. G. , Godoy-Vitorino , F. , Knight , R. & Blaser , M. J . Role of the microbiome in human development . Gut 68 , 1108 – 1114 ( 2019 ). OpenUrl Abstract / FREE Full Text 4. ↵ Bäckhed , F. et al. Dynamics and stabilization of the human gut microbiome during the first year of life . Cell Host Microbe 17 , 690 – 703 ( 2015 ). OpenUrl CrossRef PubMed 5. ↵ Abdill , R. J. , Adamowicz , E. M. & Blekhman , R . Public human microbiome data are dominated by highly developed countries . PLoS Biol . 20 , e3001536 ( 2022 ). OpenUrl CrossRef PubMed 6. ↵ Dominguez-Bello , M. G. et al. Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns . Proc. Natl. Acad. Sci. U. S. A . 107 , 11971 – 11975 ( 2010 ). OpenUrl Abstract / FREE Full Text 7. ↵ Bokulich , N. A. et al. Antibiotics, birth mode, and diet shape microbiome maturation during early life . Sci. Transl. Med . 8 , 343ra82 ( 2016 ). OpenUrl Abstract / FREE Full Text 8. ↵ Shao , Y. et al. Stunted microbiota and opportunistic pathogen colonization in caesarean-section birth . Nature 574 , 117 – 121 ( 2019 ). OpenUrl CrossRef PubMed 9. ↵ Sizear , M. I. & Rashid , M . Urgent need to address increasing caesarean section rates in lower-middle-income countries like Bangladesh . Front. Glob. Womens Health 5 , 1365504 ( 2024 ). OpenUrl 10. ↵ Khatun , H. et al. Uncovering the barriers to exclusive breastfeeding for mothers living in Dhaka’s slums: a mixed method study . Int. Breastfeed. J . 13 , 44 ( 2018 ). OpenUrl PubMed 11. Islam , M. R. et al. Prevalence and barriers to early initiation of breastfeeding among urban poor full-time readymade garments working mothers: a mixed-methods study in Bangladesh . Int. Breastfeed. J . 19 , 42 ( 2024 ). OpenUrl 12. ↵ Ahmmed , F. et al. The trend in exclusive breastfeeding practice and its association with maternal employment in Bangladesh: A multilevel analysis . Front. Public Health 10 , 988016 ( 2022 ). OpenUrl 13. ↵ Brockway , M. M. et al. Human milk bioactive components and child growth and body composition in the first 2 years: A systematic review . Adv. Nutr . 15 , 100127 ( 2024 ). OpenUrl CrossRef 14. ↵ Dinleyici , M. , Barbieur , J. , Dinleyici , E. C. & Vandenplas , Y . Functional effects of human milk oligosaccharides (HMOs) . Gut Microbes 15 , 2186115 ( 2023 ). OpenUrl 15. ↵ Bode , L . Human milk oligosaccharides: every baby needs a sugar mama . Glycobiology 22 , 1147 – 1162 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 16. ↵ Thomson , P. , Medina , D. A. & Garrido , D . Human milk oligosaccharides and infant gut bifidobacteria: Molecular strategies for their utilization . Food Microbiol . 75 , 37 – 46 ( 2018 ). OpenUrl CrossRef PubMed 17. ↵ Lewis , Z. T. et al. Maternal fucosyltransferase 2 status affects the gut bifidobacterial communities of breastfed infants . Microbiome 3 , 13 ( 2015 ). OpenUrl CrossRef PubMed 18. ↵ Reyman , M. et al. Effects of early-life antibiotics on the developing infant gut microbiome and resistome: a randomized trial . Nat. Commun . 13 , 893 ( 2022 ). OpenUrl CrossRef PubMed 19. ↵ Yassour , M. et al. Natural history of the infant gut microbiome and impact of antibiotic treatment on bacterial strain diversity and stability . Sci. Transl. Med . 8 , 343r a81 ( 2016 ). OpenUrl CrossRef PubMed 20. ↵ Luchen , C. C. et al. Impact of antibiotics on gut microbiome composition and resistome in the first years of life in low- to middle-income countries: A systematic review . PLoS Med . 20 , e1004235 ( 2023 ). OpenUrl CrossRef PubMed 21. ↵ Biswas , M. et al. Self medicated antibiotics in Bangladesh: a cross-sectional health survey conducted in the Rajshahi City . BMC Public Health 14 , 847 ( 2014 ). OpenUrl CrossRef PubMed 22. Otaigbe , I. I. & Elikwu , C. J. Drivers of inappropriate antibiotic use in low- and middle-income countries . JAC Antimicrob. Resist. 5 , dlad062 ( 2023 ). OpenUrl 23. ↵ Ndaki , P. M. et al. Drivers of inappropriate use of antibiotics among community members in low- and middle-income countries: a systematic review of qualitative studies . BMC Public Health 25 , 705 ( 2025 ). OpenUrl CrossRef PubMed 24. ↵ Prüss-Ustün , A. et al. Burden of disease from inadequate water, sanitation and hygiene for selected adverse health outcomes: An updated analysis with a focus on low- and middle-income countries . Int. J. Hyg. Environ. Health 222 , 765 – 777 ( 2019 ). OpenUrl CrossRef PubMed 25. ↵ Pell , L. G. , et al. Neonatal administration of Lactiplantibacillus plantarum ATCC 202195 with or without fructooligosaccharide in Bangladesh: a placebo-controlled randomized trial . mSphere 10 , e0103224 ( 2025 ). OpenUrl CrossRef PubMed 26. ↵ Ballard , O. & Morrow , A. L . Human milk composition: nutrients and bioactive factors . Pediatr. Clin. North Am . 60 , 49 – 74 ( 2013 ). OpenUrl CrossRef PubMed 27. ↵ Mannochio-Russo , H. et al. The microbiome diversifies long- to short-chain fatty acid-derived N-acyl lipids . Cell 0 , ( 2025 ). 28. ↵ Poulsen , K. O. et al. Dynamic changes in the human milk metabolome over 25 weeks of lactation . Front. Nutr . 9 , 917659 ( 2022 ). OpenUrl 29. ↵ Pristner , M. et al. Neuroactive metabolites and bile acids are altered in extremely premature infants with brain injury . Cell Rep. Med . 5 , 101480 ( 2024 ). OpenUrl 30. ↵ Mohanty , I. et al. The underappreciated diversity of bile acid modifications . Cell 187 , 1801 – 1818 .e20 ( 2024 ). OpenUrl CrossRef PubMed 31. ↵ Gentry , E. C. et al. Reverse metabolomics for the discovery of chemical structures from humans . Nature 626 , 419 – 426 ( 2024 ). OpenUrl CrossRef PubMed 32. ↵ Mohanty , I. et al. MS/MS mass spectrometry filtering tree for bile acid isomer annotation . bioRxiv 2025.03.04.641505 ( 2025 ) doi: 10.1101/2025.03.04.641505 . OpenUrl Abstract / FREE Full Text 33. ↵ Rimal , B. et al. Bile salt hydrolase catalyses formation of amine-conjugated bile acids . Nature 626 , 859 – 863 ( 2024 ). OpenUrl CrossRef PubMed 34. ↵ Guzior , D. V. et al. Bile salt hydrolase acyltransferase activity expands bile acid diversity . Nature 626 , 852 – 858 ( 2024 ). OpenUrl CrossRef PubMed 35. ↵ Mukhopadhya , I. & Louis , P . Gut microbiota-derived short-chain fatty acids and their role in human health and disease . Nat. Rev. Microbiol . 1 – 17 ( 2025 ). 36. ↵ Martino , C. , et al. A novel sparse compositional technique reveals microbial perturbations . mSystems 4 , ( 2019 ). 37. ↵ Orrhage , K. & Nord , C. E . Factors controlling the bacterial colonization of the intestine in breastfed infants . Acta Paediatr. Suppl . 88 , 47 – 57 ( 1999 ). OpenUrl CrossRef PubMed 38. ↵ Shenhav , L. et al. FEAST: fast expectation-maximization for microbial source tracking . Nat. Methods 16 , 627 – 632 ( 2019 ). OpenUrl CrossRef PubMed 39. ↵ Wiley , K. S. et al. Contact with caregivers is associated with composition of the infant gastrointestinal microbiome in the first 6 months of life . Am. J. Biol. Anthropol . 183 , e24858 ( 2024 ). OpenUrl 40. ↵ Lane , A. A. et al. Household composition and the infant fecal microbiome: The INSPIRE study . Am. J. Phys. Anthropol . 169 , 526 – 539 ( 2019 ). OpenUrl CrossRef PubMed 41. ↵ Ferretti , P. et al. Mother-to-infant microbial transmission from different body sites shapes the developing infant gut microbiome . Cell Host Microbe 24 , 133 – 145 .e5 ( 2018 ). OpenUrl CrossRef PubMed 42. ↵ Korpela , K. et al. Selective maternal seeding and environment shape the human gut microbiome . Genome Res . 28 , 561 – 568 ( 2018 ). OpenUrl Abstract / FREE Full Text 43. ↵ Coker , J. K. , Moyne , O. , Rodionov , D. A. & Zengler , K . Carbohydrates great and small, from dietary fiber to sialic acids: How glycans influence the gut microbiome and affect human health . Gut Microbes 13 , 1 – 18 ( 2021 ). OpenUrl CrossRef PubMed 44. ↵ Hauser , J. et al. Sialylated human milk oligosaccharides program cognitive development through a non-genomic transmission mode . Mol. Psychiatry 26 , 2854 – 2871 ( 2021 ). OpenUrl CrossRef 45. ↵ Wang , B. et al. The physiological characteristics and applications of sialic acid . Npj Sci. Food 9 , 28 ( 2025 ). OpenUrl 46. ↵ Wang , G. et al. Microbiota-derived indoles alleviate intestinal inflammation and modulate microbiome by microbial cross-feeding . Microbiome 12 , 59 ( 2024 ). OpenUrl CrossRef PubMed 47. ↵ Caminero , A. , Guzman , M. , Libertucci , J. & Lomax , A. E . The emerging roles of bacterial proteases in intestinal diseases . Gut Microbes 15 , 2181922 ( 2023 ). OpenUrl CrossRef PubMed 48. ↵ Mueller , N. T. et al. Maternal bacterial engraftment in multiple body sites of Cesarean section born neonates after vaginal seeding-a randomized controlled trial . MBio 14 , e0049123 ( 2023 ). OpenUrl 49. ↵ Lamichhane , S. et al. Trajectories of microbiome-derived bile acids in early life - insights into the progression to islet autoimmunity . medRxiv 2025.02.18.25322275 ( 2025 ) doi: 10.1101/2025.02.18.25322275 . OpenUrl Abstract / FREE Full Text 50. ↵ Vanhaecke , T. , Bretin , O. , Poirel , M. & Tap , J . Drinking water source and intake are associated with distinct gut Microbiota signatures in US and UK populations . J. Nutr . 152 , 171 – 182 ( 2022 ). OpenUrl 51. ↵ Bowyer , R. C. E. et al. Associations between UK tap water and gut microbiota composition suggest the gut microbiome as a potential mediator of health differences linked to water quality . Sci. Total Environ . 739 , 139697 ( 2020 ). OpenUrl 52. ↵ Xiang , F. et al. Comprehensive review of the expanding roles of the carnitine pool in metabolic physiology: beyond fatty acid oxidation . J. Transl. Med . 23 , 324 ( 2025 ). OpenUrl CrossRef PubMed 53. ↵ Winter , J. , Popoff , M. R. , Grimont , P. & Bokkenheuser , V. D . Clostridium orbiscindens sp. nov., a human intestinal bacterium capable of cleaving the flavonoid C-ring . Int. J. Syst. Bacteriol . 41 , 355 – 357 ( 1991 ). OpenUrl CrossRef PubMed 54. ↵ Hulme , H. et al. Microbiome-derived carnitine mimics as previously unknown mediators of gut-brain axis communication . Sci. Adv . 6 , eaax6328 ( 2020 ). OpenUrl FREE Full Text 55. ↵ Zuffa , S. et al. microbeMASST: a taxonomically informed mass spectrometry search tool for microbial metabolomics data . Nat. Microbiol . 9 , 336 – 345 ( 2024 ). OpenUrl 56. ↵ Duvallet , C. , Gibbons , S. M. , Gurry , T. , Irizarry , R. A. & Alm , E. J . Meta-analysis of gut microbiome studies identifies disease-specific and shared responses . Nat. Commun . 8 , 1784 ( 2017 ). OpenUrl CrossRef PubMed 57. ↵ Turnbaugh , P. J. et al. A core gut microbiome in obese and lean twins . Nature 457 , 480 – 484 ( 2009 ). OpenUrl CrossRef PubMed Web of Science 58. ↵ Azad , M. B. et al. Human milk oligosaccharide concentrations are associated with multiple fixed and modifiable maternal characteristics, environmental factors, and feeding practices . J. Nutr . 148 , 1733 – 1742 ( 2018 ). OpenUrl CrossRef PubMed 59. Ambalavanan , A. et al. Human milk oligosaccharides are associated with maternal genetics and respiratory health of human milk-fed children . Nat. Commun . 15 , 7735 ( 2024 ). OpenUrl 60. ↵ Asher , A. T. et al. Human milk oligosaccharide profile across lactation stages in Israeli women-A prospective observational study . Nutrients 15 , 2548 ( 2023 ). OpenUrl 61. ↵ Dührkop , K. et al. SIRIUS 4: a rapid tool for turning tandem mass spectra into metabolite structure information . Nat. Methods 16 , 299 – 302 ( 2019 ). OpenUrl CrossRef PubMed 62. ↵ Dührkop , K. et al. Systematic classification of unknown metabolites using high-resolution fragmentation mass spectra . Nat. Biotechnol . 39 , 462 – 471 ( 2021 ). OpenUrl CrossRef PubMed 63. ↵ Kunz , C. et al. Influence of gestational age, secretor, and Lewis blood group status on the oligosaccharide content of human milk . J. Pediatr. Gastroenterol. Nutr . 64 , 789 – 798 ( 2017 ). OpenUrl CrossRef PubMed 64. ↵ Totten , S. M. et al. Comprehensive profiles of human milk oligosaccharides yield highly sensitive and specific markers for determining secretor status in lactating mothers . J. Proteome Res . 11 , 6124 – 6133 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 65. ↵ Shen , W. , de Boer , J. F. , Kuipers , F. & Fu , J . New insights in amino sugar metabolism by the gut microbiome . Gut Microbes 17 , 2510462 ( 2025 ). OpenUrl 66. ↵ Wang , A. et al. Impact of milk secretor status on the fecal metabolome and microbiota of breastfed infants . Gut Microbes 15 , 2257273 ( 2023 ). OpenUrl 67. ↵ Renwick , S. et al. Modulating the developing gut microbiota with 2’-fucosyllactose and pooled human milk oligosaccharides . Microbiome 13 , 44 ( 2025 ). OpenUrl 68. ↵ Fleishman , J. S. & Kumar , S . Bile acid metabolism and signaling in health and disease: molecular mechanisms and therapeutic targets . Signal Transduct. Target. Ther . 9 , 97 ( 2024 ). OpenUrl 69. ↵ Seki , D. et al. Mutualistic interactions between Escherichia coli and Bifidobacterium bifidum enable degradation of human milk oligosaccharides in healthy infants . bioRxiv 2025.03.24.645027 ( 2025 ) doi: 10.1101/2025.03.24.645027 . OpenUrl Abstract / FREE Full Text 70. ↵ Ouyang , R. et al. Maturation of the gut metabolome during the first year of life in humans . Gut Microbes 15 , 2231596 ( 2023 ). OpenUrl CrossRef PubMed 71. ↵ Tanaka , M. et al. The association between gut microbiota development and maturation of intestinal bile acid metabolism in the first 3 y of healthy Japanese infants . Gut Microbes 11 , 205 – 216 ( 2020 ). OpenUrl CrossRef PubMed 72. ↵ Lamichhane , S. et al. Trajectories of microbiome-derived bile acids in early life – insights into the progression to islet autoimmunity . Endocrinology (including Diabetes Mellitus and Metabolic Disease) ( 2025 ). 73. ↵ Quinn , R. A. et al. Global chemical effects of the microbiome include new bile-acid conjugations . Nature 579 , 123 – 129 ( 2020 ). OpenUrl CrossRef PubMed 74. ↵ Dubrac , S. , Elentner , A. , Ebner , S. , Horejs-Hoeck , J. & Schmuth , M . Modulation of T lymphocyte function by the pregnane X receptor . J. Immunol . 184 , 2949 – 2957 ( 2010 ). OpenUrl Abstract / FREE Full Text 75. ↵ Virmani , M. A. & Cirulli , M . The role of l-carnitine in mitochondria, prevention of metabolic inflexibility and disease initiation . Int. J. Mol. Sci . 23 , 2717 ( 2022 ). OpenUrl 76. ↵ Rutkowsky , J. M. et al. Acylcarnitines activate proinflammatory signaling pathways . Am. J. Physiol. Endocrinol. Metab . 306 , E1378 – 87 ( 2014 ). OpenUrl CrossRef PubMed 77. ↵ Mahfouz , R. et al. L-Carnitine augments probenecid anti-inflammatory effect in monoiodoacetate-induced knee osteoarthritis in rats: involvement of miRNA-373/P2X7/NLRP3/NF-κB milieu . Inflammopharmacology 32 , 715 – 731 ( 2024 ). OpenUrl 78. ↵ Moossavi , S. , Miliku , K. , Sepehri , S. , Khafipour , E. & Azad , M. B . The prebiotic and probiotic properties of human milk: Implications for infant immune development and pediatric asthma . Front. Pediatr . 6 , 197 ( 2018 ). OpenUrl 79. ↵ Saima , S. , et al. Detecting Enteric pathogens in low-risk drinking water in Dhaka, Bangladesh: An assessment of the WHO water safety categories . Trop. Med. Infect. Dis . 8 , 321 ( 2023 ). OpenUrl 80. ↵ Sharif , N. et al. Antimicrobial resistant enteric bacteria are widely distributed among environmental water sources in Dhaka , Bangladesh. Npj Clean Water 8 , 1 – 15 ( 2025 ). OpenUrl 81. ↵ Xiong , S. et al. Prenatal exposure to trace elements impacts mother-infant gut microbiome, metabolome and resistome during the first year of life . Nat. Commun . 16 , 5186 ( 2025 ). OpenUrl 82. ↵ Lemons , J. M. S. et al. Enterobacteriaceae growth promotion by intestinal acylcarnitines, a biomarker of dysbiosis in inflammatory bowel disease . Cell. Mol. Gastroenterol. Hepatol . 17 , 131 – 148 ( 2024 ). OpenUrl CrossRef PubMed 83. ↵ Mihalik , S. J. et al. Increased levels of plasma acylcarnitines in obesity and type 2 diabetes and identification of a marker of glucolipotoxicity . Obesity (Silver Spring) 18 , 1695 – 1700 ( 2010 ). OpenUrl CrossRef PubMed 84. Guasch-Ferré , M. et al. Plasma acylcarnitines and risk of cardiovascular disease: effect of Mediterranean diet interventions . Am. J. Clin. Nutr . 103 , 1408 – 1416 ( 2016 ). OpenUrl Abstract / FREE Full Text 85. ↵ Issa , J. et al. Increased acylcarnitines in infant heart failure indicate fatty acid oxidation inhibition: towards therapeutic options? ESC Heart Fail . 10 , 3114 – 3122 ( 2023 ). OpenUrl 86. ↵ Andrikopoulou , M. et al. Antibiotic use without indication during delivery hospitalizations in the United States . Obstet. Gynecol . 134 , 718 – 725 ( 2019 ). OpenUrl 87. ↵ National Institute of Population Research and Training, Medical Education and Family Welfare Division, Ministry of Health and Family Welfare, Dhaka, Bangladesh, and The DHS Program, ICF, Rockville, Maryland, USA. Bangladesh Demographic and Health Survey 2022: Key Indicators Report . https://dhsprogram.com/pubs/pdf/FR386/FR386.pdf . 88. ↵ Minich , J. et al. Culture-independent meta-pangenomics enabled by long-read metagenomics reveals novel associations with pediatric undernutrition . ( 2024 ) doi: 10.2139/ssrn.4960739 . OpenUrl CrossRef 89. ↵ Anderson , M. J . Permutational multivariate analysis of variance (PERMANOVA) . Wiley StatsRef: Statistics Reference Online 1 – 15 ( 2017 ) doi: 10.1002/9781118445112.stat07841 . OpenUrl CrossRef 90. ↵ Ernst , M. et al. Gestational age-dependent development of the neonatal metabolome . Pediatr. Res . 89 , 1396 – 1404 ( 2021 ). OpenUrl 91. ↵ Ma , J. et al. Comparison of gut microbiota in exclusively breast-fed and formula-fed babies: a study of 91 term infants . Sci. Rep . 10 , 15792 ( 2020 ). OpenUrl CrossRef PubMed 92. ↵ Chalifour , B. et al. The potential role of early life feeding patterns in shaping the infant fecal metabolome: implications for neurodevelopmental outcomes . NPJ Metab. Health Dis . 1 , 2 ( 2023 ). OpenUrl 93. ↵ Mansell , T. et al. The newborn metabolome: associations with gestational diabetes, sex, gestation, birth mode, and birth weight . Pediatr. Res . 91 , 1864 – 1873 ( 2022 ). OpenUrl 94. ↵ Kwak , S. et al. Sociobiome - Individual and neighborhood socioeconomic status influence the gut microbiome in a multi-ethnic population in the US . NPJ Biofilms Microbiomes 10 , 19 ( 2024 ). OpenUrl 95. ↵ Azad , M. B. et al. Infant gut microbiota and the hygiene hypothesis of allergic disease: impact of household pets and siblings on microbiota composition and diversity . Allergy Asthma Clin. Immunol . 9 , 15 ( 2013 ). OpenUrl CrossRef PubMed 96. ↵ Brennan , C. , et al. Clearing the plate: a strategic approach to mitigate well-to-well contamination in large-scale microbiome studies . mSystems 9 , e0098524 ( 2024 ). OpenUrl CrossRef PubMed 97. ↵ Zuffa , S. et al. Human untargeted metabolomics in high-throughput gut microbiome research: Ethanol vs methanol . Anal. Chem . 97 , 4945 – 4953 ( 2025 ). OpenUrl 98. ↵ Villaseñor , A. et al. Breast milk metabolome characterization in a single-phase extraction, multiplatform analytical approach . Anal. Chem . 86 , 8245 – 8252 ( 2014 ). OpenUrl 99. ↵ Chambers , M. C. et al. A cross-platform toolkit for mass spectrometry and proteomics . Nat. Biotechnol . 30 , 918 – 920 ( 2012 ). OpenUrl CrossRef PubMed 100. ↵ Schmid , R. et al. Integrative analysis of multimodal mass spectrometry data in MZmine 3 . Nat. Biotechnol . 41 , 447 – 449 ( 2023 ). OpenUrl CrossRef PubMed 101. ↵ Nothias , L.-F. et al. Feature-based molecular networking in the GNPS analysis environment . Nat. Methods 17 , 905 – 908 ( 2020 ). OpenUrl CrossRef PubMed 102. ↵ Wang , M. et al. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking . Nat. Biotechnol . 34 , 828 – 837 ( 2016 ). OpenUrl CrossRef PubMed 103. ↵ Shannon , P. et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks . Genome Res . 13 , 2498 – 2504 ( 2003 ). OpenUrl Abstract / FREE Full Text 104. ↵ Mildau , K. et al. Homologue series detection and management in LC-MS data with homologueDiscoverer . Bioinformatics 38 , 5139 – 5140 ( 2022 ). OpenUrl PubMed 105. ↵ Damiani , T. et al. A universal language for finding mass spectrometry data patterns . Nat. Methods 1 – 8 ( 2025 ). 106. ↵ Conway , J. R. , Lex , A. & Gehlenborg , N . UpSetR: an R package for the visualization of intersecting sets and their properties . Bioinformatics 33 , 2938 – 2940 ( 2017 ). OpenUrl CrossRef PubMed 107. ↵ Shaffer , J. P. et al. A comparison of DNA/RNA extraction protocols for high-throughput sequencing of microbial communities . Biotechniques 70 , 149 – 159 ( 2021 ). OpenUrl CrossRef PubMed 108. ↵ Sanders , J. G. et al. Optimizing sequencing protocols for leaderboard metagenomics by combining long and short reads . Genome Biol . 20 , 226 ( 2019 ). OpenUrl CrossRef PubMed 109. ↵ Brennan , C. , et al. Maximizing the potential of high-throughput next-generation sequencing through precise normalization based on read count distribution . mSystems 8 , e0000623 ( 2023 ). OpenUrl CrossRef PubMed 110. ↵ Armstrong , G. , et al. Swapping metagenomics preprocessing pipeline components offers speed and sensitivity increases . mSystems 7 , e0137821 ( 2022 ). OpenUrl CrossRef PubMed 111. ↵ Chen , S. , Zhou , Y. , Chen , Y. & Gu , J. fastp: an ultra-fast all-in-one FASTQ preprocessor . Bioinformatics 34 , i884 – i890 ( 2018 ). OpenUrl CrossRef PubMed 112. ↵ Guccione , C. et al. Incomplete human reference genomes can drive false sex biases and expose patient-identifying information in metagenomic data . Nat. Commun . 16 , 825 ( 2025 ). OpenUrl 113. ↵ Li , H . Minimap2: pairwise alignment for nucleotide sequences . Bioinformatics 34 , 3094 – 3100 ( 2018 ). OpenUrl CrossRef PubMed 114. ↵ Liao , W.-W. et al. A draft human pangenome reference . Nature 617 , 312 – 324 ( 2023 ). OpenUrl CrossRef PubMed 115. ↵ Zakeri , M. , Brown , N. K. , Ahmed , O. Y. , Gagie , T. & Langmead , B . Movi: A fast and cache-efficient full-text pangenome index . iScience 27 , 111464 ( 2024 ). OpenUrl 116. ↵ Gonzalez , A. et al. Qiita: rapid, web-enabled microbiome meta-analysis . Nat Methods 15 , 796 – 798 ( 2018 ). OpenUrl CrossRef PubMed 117. ↵ Zhu , Q. , et al. Phylogeny-Aware Analysis of Metagenome Community Ecology Based on Matched Reference Genomes while Bypassing Taxonomy . mSystems 7 , e0016722 ( 2022 ). OpenUrl CrossRef PubMed 118. ↵ Hillmann , B. et al. SHOGUN: a modular, accurate and scalable framework for microbiome quantification . Bioinformatics 36 , 4088 – 4090 ( 2020 ). OpenUrl CrossRef PubMed 119. ↵ Zhu , Q. et al. Phylogenomics of 10,575 genomes reveals evolutionary proximity between domains Bacteria and Archaea . Nat Commun 10 , 5477 ( 2019 ). OpenUrl CrossRef PubMed 120. ↵ Langmead , B. & Salzberg , S. L . Fast gapped-read alignment with Bowtie 2 . Nat Methods 9 , 357 – 359 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 121. ↵ McDonald , D. et al. The Biological Observation Matrix (BIOM) format or: how I learned to stop worrying and love the ome-ome . Gigascience 1 , 7 ( 2012 ). OpenUrl CrossRef PubMed 122. ↵ Bolyen , E. et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2 . Nat Biotechnol 37 , 852 – 857 ( 2019 ). OpenUrl CrossRef PubMed 123. ↵ Austin , G. I. et al. Contamination source modeling with SCRuB improves cancer phenotype prediction from microbiome data . Nat. Biotechnol . 41 , 1820 – 1828 ( 2023 ). OpenUrl CrossRef PubMed 124. ↵ Faith , D. P. Conservation evaluation and phylogenetic diversity . Biol. Conserv. 61 , 1 – 10 ( 1992 ). OpenUrl 125. ↵ Shannon , C. E. , Weaver , W. & Wiener , N . The mathematical theory of communication . Phys. Today ( 1950 ). 126. ↵ Lozupone , C. , Lladser , M. E. , Knights , D. , Stombaugh , J. & Knight , R . UniFrac: an effective distance metric for microbial community comparison . ISME J 5 , 169 – 172 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 127. ↵ Vázquez-Baeza , Y. , Pirrung , M. , Gonzalez , A. & Knight , R . EMPeror: a tool for visualizing high-throughput microbial community data . Gigascience 2 , 16 ( 2013 ). OpenUrl CrossRef PubMed 128. ↵ Shi , P. et al. TEMPTED: time-informed dimensionality reduction for longitudinal microbiome studies . Genome Biol . 25 , 317 ( 2024 ). OpenUrl CrossRef PubMed 129. ↵ Lin , H. & Peddada , S. D . Multigroup analysis of compositions of microbiomes with covariate adjustments and repeated measures . Nat. Methods 21 , 83 – 91 ( 2024 ). OpenUrl CrossRef PubMed 130. ↵ Fedarko , M. W. , et al. Visualizing ’omic feature rankings and log-ratios using Qurro . NAR Genom. Bioinform. 2 , lqaa023 ( 2020 ). OpenUrl 131. ↵ Song , S. J. et al. Naturalization of the microbiota developmental trajectory of Cesarean-born neonates after vaginal seeding . Med 2 , 951 – 964 .e5 ( 2021 ). OpenUrl 132. ↵ UniProt Consortium . UniProt: a worldwide hub of protein knowledge . Nucleic Acids Res . 47 , D506 – D515 ( 2019 ). OpenUrl CrossRef PubMed 133. ↵ Zhou , Y. , Liu , Y.-X. & Li , X . USEARCH 12: Open-source software for sequencing analysis in bioinformatics and microbiome . Imeta 3 , e236 ( 2024 ). OpenUrl 134. ↵ Buchfink , B. , Reuter , K. & Drost , H.-G . Sensitive protein alignments at tree-of-life scale using DIAMOND . Nat. Methods 18 , 366 – 368 ( 2021 ). OpenUrl CrossRef PubMed 135. ↵ Berger , P. K. et al. Stability of human-milk oligosaccharide concentrations over 1 week of lactation and over 6 hours following a standard meal . J. Nutr . 152 , 2727 – 2733 ( 2023 ). OpenUrl CrossRef PubMed 136. ↵ Berger , P. K. et al. Human milk oligosaccharide 2’-fucosyllactose links feedings at 1 month to cognitive development at 24 months in infants of normal and overweight mothers . PLoS One 15 , e0228323 ( 2020 ). OpenUrl CrossRef PubMed 137. Filmer , D. & Pritchett , L. H . Estimating wealth effects without expenditure data--or tears: an application to educational enrollments in states of India . Demography 38 , 115 – 132 ( 2001 ). OpenUrl CrossRef PubMed Web of Science View the discussion thread. Back to top Previous Next Posted July 24, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Environmental and Maternal Imprints on Infant Gut Metabolic Programming Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Environmental and Maternal Imprints on Infant Gut Metabolic Programming Kine Eide Kvitne , Celeste Allaband , Jennifer C. Onuora , Daniela Perry , Simone Zuffa , Lucas Patel , Vincent Charron-Lamoureux , Ipsita Mohanty , Kristija Sejane , Abubaker Patan , Abdullah Al Mahmud , Tahmeed Ahmed , Diego G. Bassani , Antonio González , Davidson H. Hamer , Rashidul Haque , Benjamin Ho , Md Iqbal Hossain , Mohammad Shahidul Islam , Daniel McDonald , Lisa G. Pell , Huma Qamar , Daniel E. Roth , Samir Saha , Prakesh S. Shah , Md Muniruzzaman Siddiqui , Shafiqul Alam Sarker , Shamima Sultana , Sydney Thomas , Lindsey A. Burnett , Shirley M. Tsunoda , Lars Bode , Pieter C. Dorrestein , Rob Knight bioRxiv 2025.07.24.666662; doi: https://doi.org/10.1101/2025.07.24.666662 Share This Article: Copy Citation Tools Environmental and Maternal Imprints on Infant Gut Metabolic Programming Kine Eide Kvitne , Celeste Allaband , Jennifer C. Onuora , Daniela Perry , Simone Zuffa , Lucas Patel , Vincent Charron-Lamoureux , Ipsita Mohanty , Kristija Sejane , Abubaker Patan , Abdullah Al Mahmud , Tahmeed Ahmed , Diego G. Bassani , Antonio González , Davidson H. Hamer , Rashidul Haque , Benjamin Ho , Md Iqbal Hossain , Mohammad Shahidul Islam , Daniel McDonald , Lisa G. Pell , Huma Qamar , Daniel E. Roth , Samir Saha , Prakesh S. Shah , Md Muniruzzaman Siddiqui , Shafiqul Alam Sarker , Shamima Sultana , Sydney Thomas , Lindsey A. Burnett , Shirley M. Tsunoda , Lars Bode , Pieter C. Dorrestein , Rob Knight bioRxiv 2025.07.24.666662; doi: https://doi.org/10.1101/2025.07.24.666662 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Microbiology Subject Areas All Articles Animal Behavior and Cognition (7635) Biochemistry (17690) Bioengineering (13892) Bioinformatics (41936) Biophysics (21451) Cancer Biology (18588) Cell Biology (25499) Clinical Trials (138) Developmental Biology (13378) Ecology (19899) Epidemiology (2067) Evolutionary Biology (24320) Genetics (15609) Genomics (22506) Immunology (17736) Microbiology (40394) Molecular Biology (17181) Neuroscience (88603) Paleontology (666) Pathology (2832) Pharmacology and Toxicology (4824) Physiology (7641) Plant Biology (15152) Scientific Communication and Education (2045) Synthetic Biology (4294) Systems Biology (9825) Zoology (2271)