Intestinal flora and inflammatory bowel disease: Causal relationships and predictive models

Area of Science:

• Genetics
• Microbiology
• Computational Biology

Background:

• Inflammatory bowel disease (IBD), encompassing Crohn's disease and ulcerative colitis, is strongly influenced by the gut microbiome.
• Understanding the complex genetic and microbial interactions is vital for effective IBD prediction and treatment strategies.

Purpose of the Study:

• To elucidate the causal relationships between intestinal flora and IBD using an integrated multi-omics approach.
• To develop and validate machine learning models for predicting IBD onset and identifying potential therapeutic targets.

Main Methods:

• Employed Mendelian randomization (MR), transcriptomic analysis, and machine learning (LASSO regression, six predictive models).
• Integrated data from MiBioGen Consortium and GWAS datasets, mapping SNPs to genes and identifying differentially expressed genes (DEGs).
• Analyzed immune infiltration patterns and utilized soft voting for enhanced model reliability.

Main Results:

• Identified 25 gut microbiota classifications causally linked to IBD through MR analysis.
• Highlighted 24 significant genes and validated their causal relationship with IBD via drug target MR and colocalization.
• Revealed key pathways (PI3K-Akt, epithelial-mesenchymal transition) and distinct immune infiltration patterns.
• Demonstrated high predictive accuracy for IBD using machine learning models.

Conclusions:

• This study establishes a robust framework for predicting IBD by integrating multi-omics data and advanced computational methods.
• The findings provide novel insights into IBD pathogenesis and identify potential therapeutic targets by clarifying the gut microbiome's causal role.