Host-microbiome-dietary interactions play crucial roles in regulating human health, yet their direct functional assessment remains challenging. We adopted metagenome-informed metaproteomics (MIM), in mice and humans, to non-invasively explore species-level microbiome-host interactions during commensal and pathogen colonization, nutritional modification, and antibiotic-induced perturbation. Simultaneously, fecal MIM accurately characterized the nutritional exposure landscape in multiple clinical and dietary contexts. Implementation of MIM in murine auto-inflammation and in human inflammatory bowel disease (IBD) characterized a "compositional dysbiosis" and a concomitant species-specific "functional dysbiosis" driven by suppressed commensal responses to inflammatory host signals. Microbiome transfers unraveled early-onset kinetics of these host-commensal cross-responsive patterns, while predictive analyses identified candidate fecal host-microbiome IBD biomarker protein pairs outperforming S100A8/S100A9 (calprotectin). Importantly, a simultaneous fecal nutritional MIM assessment enabled the determination of IBD-related consumption patterns, dietary treatment compliance, and small intestinal digestive aberrations. Collectively, a parallelized dietary-bacterial-host MIM assessment functionally uncovers trans-kingdom interactomes shaping gastrointestinal ecology while offering personalized diagnostic and therapeutic insights into microbiome-associated disease.