Identifying inflammatory bowel disease subtypes: a comprehensive exploration of transcriptomic data and machine...

Area of Science:

• Genomics
• Molecular Biology
• Immunology

Background:

• Inflammatory bowel disease (IBD), comprising Crohn's disease (CD) and ulcerative colitis (UC), is characterized by chronic gastrointestinal inflammation and immune dysregulation.
• Current molecular subtyping of IBD faces challenges in consistency across datasets, hindering personalized treatment.
• There is a significant need for robust IBD subtypes that capture disease heterogeneity and predict clinical outcomes.

Purpose of the Study:

• To identify transcriptomic subtypes in both ulcerative colitis (UC) and Crohn's disease (CD) using a large-scale RNA sequencing dataset.
• To analyze molecular characteristics and clinical relevance of identified transcriptomic subtypes.
• To support a stratified approach for personalized IBD diagnosis and therapy.

Main Methods:

• Analysis of RNA sequencing data from four prospective cohorts (GSE193677, GSE186507, GSE137344, GSE235236) comprising 2490 adult IBD patients.
• Application of k-means clustering to identify transcriptomic subtypes in UC and CD patient samples.
• Gene set enrichment, network analyses, and statistical tests (Chi-square, ANOVA) to assess molecular features and clinical associations.

Main Results:

• Three distinct transcriptomic subtypes were identified in both UC and CD using k-means clustering.
• UC subtypes were enriched for RNA processing/DNA repair, autophagy/stress responses, or cytoskeletal organization.
• CD subtypes showed variations in cytoskeletal remodeling, protein synthesis, stress, and translation pathways, with Cluster 3 associated with moderate-to-severe disease activity and Cluster 1 with mild disease.

Conclusions:

• The study successfully defined three transcriptomic subtypes for both UC and CD, each with unique molecular signatures.
• These subtypes demonstrate significant clinical relevance, particularly in relation to disease severity.
• The findings advocate for a personalized, stratified approach to IBD management based on molecular subtypes.