Characterisation of Genetically-Derived Patient Clusters Highlights Genetic Heterogeneity of Inflammatory Bowel Disease

Inflammatory Bowel Disease (IBD) is a complex immune-mediated condition with a significant genetic component. Researchers have developed a gene-level disease burden metric called GenePy, which integrates clinical and genetic data to aid in patient stratification for personalized care using machine learning methods.

To study the genetic heterogeneity of IBD, a team of researchers curated consensus IBD pathway genes and computed GenePy scores from whole exome sequencing data for each gene in the identified pathways. They then applied a K-Means algorithm to the GenePy scores, optimizing the clustering results using quality and stability metrics. This process was repeated in three independent IBD cohorts, including SPARC IBD, MSCCR IBD, and Southampton IBD.

Upon analysis, the researchers found a strong association between the cluster assignments and various phenotypic features across all cohorts. This association was particularly notable with NOD signalling genes, which were found to drive IBD subtype, behavior, and surgical outcomes. The analysis also revealed additional phenotypic enrichment in clusters driven by NLRP1, as well as nominally significant enrichment for disease behavior linked to BIRC2 in some cohorts.

Overall, this study highlights the genetic heterogeneity of IBD and identifies common genetic architectures across different IBD cohorts. It further strengthens the case for NOD2 as a monogenic driver of disease with strong phenotypic enrichment, as well as other putative genetic drivers, such as NLRP1.