The gut microbiome can provide valuable insights into host pathophysiological changes associated with digestive system diseases. To date, meaningful associations have been reported in gastrointestinal disorders and hepatic and pancreatobiliary diseases. Microbiome-based analyses currently have clear limitations as standalone diagnostic tools for specific diseases owing to substantial inter-individual variability, methodological heterogeneity, and constraints in functional interpretation. Diagnostic interpretation has thus shifted from assessing changes in individual microbial taxa to identifying patterns of microbial communities with associated functional characteristics. In this context, the concepts of disease-associated core microbiome or functional core microbiome offer a useful framework for understanding shared microbial functional features related to digestive diseases. Machine learning approaches, including deep learning-based methods, enable integrative analysis of high-dimensional microbiome data and provide opportunities to explore disease-related patterns that are difficult to capture using conventional analytical techniques.
Kihyun Ryu (2026) studied this question.