Colorectal cancer (CRC) is a major cause of cancer morbidity and mortality, and increasing evidence indicates that the gut microbiome contributes to CRC development and progression. However, most microbiome studies remain descriptive and species-centred, while the microbial proteins, metabolites and molecular mechanisms responsible for host effects are still poorly understood. The overall aim of this project is to identify gut microbial and metabolic mechanisms associated with colorectal cancer by integrating large-scale metagenomic, proteomic and metabolomic data. We will analyse multiple independent human gut metagenomic cohorts to identify reproducible microbial features associated with CRC while accounting for inter-individual, cohort and technical variation. In addition to conventional taxonomic and pathway-based analyses, we will develop a protein-centric representation of the gut microbiome by constructing large microbial protein catalogues, performing sequence clustering and functional annotation, and applying protein language models to derive annotation-independent representations of microbial proteins. These analyses will be integrated with metabolomic and proteomic data to identify microbial proteins, metabolites and molecular pathways that may mediate interactions between the microbiome and host cells. Particular emphasis will be placed on distinguishing robust disease-associated signals from cohort-specific variation and on prioritizing candidate host–microbe and protein–metabolite interactions for mechanistic investigation. The project builds on our previous work in host–bacteria interaction proteomics and large-scale metagenomic analysis. The resulting framework will connect microbial community composition, microbial protein function, metabolites and host molecular responses at multiple biological levels. Ultimately, the project aims to identify reproducible microbial molecular mechanisms that may contribute to colorectal cancer and provide candidate biomarkers or targets for future experimental and therapeutic studies.