Fig. 3

Multi-omics identifies features both common and unique features of ileal- and colonic-isolated diseases. A Enrichr-generated gene ontology enrichment graph using host proteome features significantly greater in ICD compared to CCD, sorted by Enrichr combination score [23]. B Volcano plots comparing major CD subtypes and the abundance of bile acid labeled metabolites. Significance cutoff = FDR < 0.05 and |log2 fold-change| > 0.5. Bile acid types denoted by specific colors. ICD n = 31, CCD n = 31, ICCD n = 11. C Heatmap of Gammaproteobacteria feature abundances (x-axis) in ICD (n = 31) and CCD (n = 24) subjects. Color key at bottom denotes feature-set origin. D Bar plot comparing Spearman correlation values (Proteobacteria-SES-CD, only significant values selected, see) between ICD (14 features) and CCD (17 features). **** signifies p < 0.0001, Welch’s t test. E Heatmap of Lachnospiraceae mean abundances for each CD subtype. Values scaled column-wise for visualization. F Bar plot comparing Spearman correlation values (Lachnospiraceae-SES-CD, only significant features [p < 0.05] selected, CCD = 88, ICD = 45, ICCD 56) between major CD subtypes. *** signifies p < 0.001, Welch’s t test. G Heatmap of Faecalibacterium feature abundance (y-axis, all feature sets included) among CD subtypes (x-axis). Feature abundances are row-scaled for emphasis. H Plot of Faecalibacterium-bile acids correlations, colored by subtype and split by location. Only significant features used in comparison