DOP79 Alternative polygenic forms of Inflammatory Bowel Disease are not captured by current generations of polygenic risk scores

Mateos, B.(1)*;Gaite, A.(1);Rodríguez-Lago, I.(2);M. Marigorta, U.(1);

(1)cic bioGUNE, Integrative Genomics, Bilbao, Spain;(2)Galdakao Hospital, Gastroenterology, Bilbao, Spain;

Background

Polygenic risk scores (PRSs) have emerged as a promising tool to ascertain individuals at high risk of developing IBD. However, the sensitivity of current PRSs is limited, and will likely remain so in the immediate future. The default expectation is that patients not captured by PRSs must be enriched for rare variants with strong effects and/or environmental risk factors. We set out to investigate an alternative scenario, namely that combinations of common variants in non-GWAS genes conform population of patients that are missed by large genetic studies for the main forms of the disease.

Methods

MAGMA was used to identify genes from the International Inflammatory Bowel Disease Genetics Consortium (IIBDGC) GWAS. We selected a sort of monogenic IBD genes1 which were significant for IBD (pval ≈ 10-4) but do not overlap with polygenic IBD genes. These genes were compared using DESeq tool on two transcriptomic databases: RISK (Crohn’s disease (CD) onset vs non-IBD) and PROTECT (ulcerative colitis (UC) onset vs 52 weeks). A second DESeq analysis was performed between patients with high against low expression of these 13 genes. We explored these genes using in-silico predicted expression based on eQTL in 9 different tissues for UK Biobank (UKBB) population, comparing UKBB individuals according to disease and polygenic risk score status.

Results

The starting point based on MAGMA analysis led to the detection of 13 monogenic IBD genes which may partially participate in polygenic forms of the disease. The transcriptomic analyses, based on stratification according to the levels of these 13 genes, implied that these genes participate in regulatory programmes that are particularly heterogeneous and variable among cases. The detected differentially expressed genes are associated with immune processes and receptors, inflammation transporters, and digestive problems, as well as skin development, humoral response, extracellular matrix, and mitochondrial component.  The TWAS-based estimation analyses hint that low-PRS patients are indeed enriched for common variants associated with higher pathogenic levels at the 13 selected genes.

Conclusion

Although polygenic risk scores may play a valuable role in patient management, our study hints at the presence of patients with common polygenic forms that are not captured by current genetic risk estimators.

1. Chrissy B. et al, 2022. Gastroenterology.