Improved integrative framework combining association data with gene expression features to prioritize Crohn's disease...

Area of Science:

• Genetics and Genomics
• Immunology
• Computational Biology

Background:

• Genome-wide association studies (GWAS) have identified numerous loci for Crohn's disease (CD), but pinpointing causative genes remains difficult.
• Standard GWAS thresholds balance false positives and negatives, potentially missing relevant genes.
• Understanding the genetic underpinnings of CD requires integrating diverse data types beyond simple association signals.

Purpose of the Study:

• To develop and validate an integrative computational model for prioritizing genes associated with Crohn's disease (CD) pathogenicity.
• To improve the identification of CD-related genes by incorporating gene expression and autoimmune disease association data.
• To provide a standardized framework for prioritizing genes, particularly in complex genomic regions and for rare variants.

Main Methods:

• Developed an integrative model using a Crohn's disease (CD) reference gene set, incorporating gene-based association data from CD and autoimmune (AI) diseases, and gene expression features (eQTLs, epigenetic markers, intestinal expression).
• Evaluated the model's performance using statistical validation, gene ontology enrichment, differential expression analysis between M1 and M2 macrophages, and validation against genes causing monogenic inflammatory bowel disease.
• Analyzed 140 high-confidence CD regions to assess gene prioritization scores and identify genes unlikely to contribute to CD pathogenesis.

Main Results:

• The integrative model significantly outperformed gene-based association signals alone in identifying CD-related genes.
• Association with autoimmune (AI) diseases emerged as a strong predictor, underscoring shared inflammatory mechanisms, notably the interferon-γ pathway.
• Within 140 CD regions, a substantial number of genes (598/1328) received low prioritization scores, suggesting they are less likely involved in CD pathogenesis.
• Complex regions like the IBD5 locus showed multiple genes with high integrative scores due to extensive linkage disequilibrium.

Conclusions:

• The developed integrative model offers a standardized and effective method for prioritizing potential Crohn's disease (CD) genes, applicable to both highly and nominally significant loci.
• This approach enhances the identification of CD-related genes by leveraging multi-omics data and cross-disease associations.
• The model shows particular promise for prioritizing rare or private missense variants where traditional GWAS may lack sufficient evidence.