Fig. 2: Baseline bile acid profile associated with POIT efficacy.
From: Gut microbial bile and amino acid metabolism associate with peanut oral immunotherapy failure
a Association between untargeted metabolomics module (UMM) eigenvectors and POIT outcomes (two-sided ANOVA, adjusted for age). b Ordination of baseline secondary BA metabolites (n = 43, R2 = 0.10; P < 0.015, adjusted for age), PERMANOVA analyses (two-sided) based on Euclidean distance matrix. c Difference in Module Eigenvectors (ME), which were determined based on WGCNA analyses (see the “Methods” section) and represents a measure of the joint abundance profile of a specific module, of UMM10. d UMM15, and e UMM4 between POIT-outcome groups. Two-sided Wilcoxon signed-rank test (n = 129; D+R+ = 33, D+R− = 78, D−R− = 18). Data are presented as mean values ± SEM. Boxplots show the median (center line), 25th and 75th percentiles (box bounds), and whiskers extend to values within 1.5× the interquartile range. f Baseline Z-scores of each BA-related metabolite of UMM10. g UMM15 and h UMM4 in each POIT-outcomes (n = 43; D+R+ = 11, D+R− = 26, D−R− = 6). Blue colors represent low z-scores thus low abundance and red colors represent high z-score and higher abundance. i POIT responders exhibit an increased copy number of the EC.1.1.1.392 enzyme compared to non-responders at baseline (P = 0.024, Two-sided Wilcoxon signed-rank test; Remission, Yes = 16, No = 44). Data are presented as mean values ± SEM. Boxplots show the median (center line), 25th and 75th percentiles (box bounds), and whiskers extend to values within 1.5× the interquartile range. j Copy number of EC.1.1.1.392 enzyme negatively correlates with lithocholate abundance at baseline samples. Two-sided Pearson correlation (n = 32, r = −0.35, p = 0.041. Adjusted for age). The shaded area around the line represents the 95% confidence interval for the fitted regression line. k Bifidobacterium breve and Ruminococcus gnavus encode significantly higher copy numbers of secondary BA production enzymes in POIT responders (n = 16) compared to non-responders (n = 44). Data were filtered to retain enzymes with log2- fold changes exceeding ±0.5. Asterisk “*” represents P < 0.05, and double Asterisk “**” represents P.FDR < 0.05. Log2 FC (cpm) represents the Log2 fold change of copies per million between remission and no-remission groups (two-sided linear mixed effect models). Red color represents an increased Log2 FC (cpm) in the remission group. Exact P values are presented in Supplementary Data 12. l The mean relative abundances of R. gnavus (P = 0.042) and B. breve (P = 0.027) are significantly enriched in POIT responders (n = 16) compared to non-responders (n = 44). Data are presented as mean values ± SEM. Boxplots show the median (center line), 25th and 75th percentiles (box bounds), and whiskers extend to values within 1.5× the interquartile range. Statistical comparisons were performed using the two-sided Wilcoxon rank-sum test. Source data are provided as a Source Data file.
