File Name : figure_s1.pdf

Caption : major dietary taxa dna detection patterns do not align with rusf administration. clr-normalized dna relative abundances of top 10 taxa driving dietary composition as identified. by pca biplot compared across all timepoints. the 10-month timepoint when rusf was administered is highlighted in orange. box plots show median and interquartile ranges, while individual dots represent samples from each subject. wilcoxon signed-rank test for 3-6 months vs 9 months and 10 months vs 14 months and wilcoxon rank-sum test for all other timepoint comparisons were conducted. significance results of statistical testing are marked. significance values in bold remain significant after benjamini-hochberg multiple testing correction (including chickpea and tea timepoint comparison tests).

File Name : figure_s2_revised.pdf

Caption : fig. s2|health status does not impact dietary composition and chickpea dna detection is not impacted by socioeconomic or other dietary factors. (a) principal component analysis (pca) of plant dietary composition with biplot indicating relative contributions of top 10 dietary components colored by health status (healthy controls vs malnourished samples) and clusters denoted by ellipses. health status compared along pc1 and pc2 values using linear mixed effect models with age, gender, and pcr batch as fixed effects and sample id as random effects. no significant impact detected when accounting for age. (b) chickpea clr-normalized relative abundances at the 10-month timepoint compared across family size, food insecurity status, wealth quintile of surveyed households using wilcoxon rank-sum test. box plots show median and interquartile ranges, while individual dots represent samples from each subject. no significant difference detected.  (c) chickpea clr-normalized relative abundances compared at the 10-month timepoint against dietary factors such as traditional and genomic dietary diversity as well as the first two principal components from a plant dietary composition pca excluding chickpea (each pc approximating a broader dietary pattern). linear mixed effects model with participant id as random effect and pcr batch, and gender as fixed effects. shaded areas represent 95% confidence intervals. no significant relationship detected. (d) tea and water buffalo dna are partially but not consistently co-detected in infant stool. scatter plot of clr-normalized tea (camellia) and water buffalo (bubalus bubalis) dna across all samples and timepoints. colors indicate infant age. tea dna abundance showed a modest positive association with milk dna (p = 0.012), but many samples contained tea dna without detectable water buffalo dna.

File Name : figure_s3.pdf

Caption : dietary dna composition in stool samples: heatmap of vertebrate taxa abundances and presence counts of common food taxa in healthy controls. (a) clr-normalized relative abundance values of  cow and goat dna detected in stool samples compared across all timepoints. box plots show median and interquartile ranges, while individual dots represent samples from each subject. wilcoxon signed-rank test for 3-6 months vs 9 months and 10 months vs 14 months and wilcoxon rank-sum test for all other timepoint comparisons were conducted. significance results of statistical testing are marked. (b) heatmap of relative abundances of vertebrate taxa dna post-filtration of human dna and taxa with less than 5 total reads. low-frequency detection of likely environmental contaminants (e.g. dog, indian mouse) may reflect poor sanitation or food contamination. (c) bar plot showing the presence counts of top 25 food (plant and vertebrate combined) taxa detected in participant stool samples in healthy controls (>12 months).

File Name : table s1.docx

Caption : top taxa identified as drivers of pc1 and pc2 in dietary principal components analysis