We describe a data-driven computational approach for quantifying the microbial functional pathways in the gastrointestinal tract, based on an analysis of stool metatranscriptomic data within a large and diverse adult human population (n = 66,545). We develop a suite of eight gut microbial pathway scores, each of which represents the activity of a well-defined set of microbial functional features relevant to known gut biochemical activities. We use a normative approach within a subpopulation (n = 9,350) to determine the optimal and non-optimal levels of activity of these functional pathway scores. We hypothesize that non-optimal scores are associated with irritable bowel syndrome (IBS) and its subtypes (i.e., IBS-Constipation, IBS-Diarrhea, IBS-Mixed Type). We show that non-optimal scores within these microbial functional pathways are associated with higher odds of IBS or its subtypes within an independent cohort (n = 57,195). Given these findings, these gut pathway scores can be used to deliver meaningful health insights from simple, non-invasive stool samples, as potential molecular endpoints to measure the efficacy of practical interventions, and develop data-driven personalized algorithms aimed at providing nutritional recommendations, diagnostics, and treatments for IBS as well as related conditions.