Fig. 6: Meta-analysis of CAZymes in metagenomes from patients with CRC and tumour-free controls.
From: Cayman enables large-scale analysis of gut microbiome carbohydrate-active enzyme repertoires
a, Heat map with univariate associations per CRC study. Asterisk signifies per-study FDR-corrected significance (BH-adjusted P value <0.05). The top and bottom 15 families were selected on the basis of the order of CAZyme families from the meta-analysis in b (all having a BH-adjusted P value <0.05). P values for differential CAZyme abundances were calculated using linear models with disease status as predictor. Data from the references listed in Supplementary Table 5. b, Meta-analysis of CAZyme enrichments over all studies using LMMs where CAZyme abundances were predicted from disease status as in a but additionally with study as a random effect. Subsequently, all CAZyme families were subjected to GSEA to identify overrepresented substrate groups. c, GSEA analysis at different hierarchical substrate levels to identify overrepresented substrate groups differentially targeted between CRC and control metagenomes. NES, normalized enrichment score. d, Dot plot depicting CRC-enriched CBMs co-occurring with catabolic CAZyme families within the same gene. Co-occurrences were calculated using a modified Jaccard index on bacterial genomic data (Methods). Dot size corresponds to the strength of association, and colour indicates substrate annotation of the enzymatic domain.
