[New publication] PredCMB: predicting changes in microbial metabolites based on the gene–metabolite network analysis of shotgun metagenome data

Our new paper has been published in Bioinformatics.

The paper introduces PredCMB, a novel computational method for predicting changes in individual microbial metabolites based on shotgun metagenome data and enzymatic gene–metabolite networks. Unlike existing approaches that infer pathway activities, PredCMB directly estimates metabolite-level alterations by analyzing differential enzymatic gene abundance and reaction abundance weights. The method was validated using two publicly available datasets from inflammatory bowel disease (IBD) and gastrectomy cohorts, showing that PredCMB outperformed previous methods by achieving stronger correlations between predicted and experimentally measured metabolite changes. PredCMB successfully identified key metabolite classes that exhibited major alterations, providing deeper insights into microbial metabolic dynamics. The tool has potential applications in refining metabolomics experiments and assessing microbial contributions to disease pathogenesis, offering a metagenome-based approach without requiring extensive training datasets or complex metabolic models.

Jungyong Ji, Sungwon Jung. "PredCMB: predicting changes in microbial metabolites based on the gene–metabolite network analysis of shotgun metagenome data" Bioinformatics, 41(1):btaf020, 2025.