P097 Various regulatory single nucleotide polymorphisms regulate similar biological pathways in Crohn’s disease

Abstract

Abstract Background Single nucleotide polymorphisms (SNP) associated with Crohn’s disease (CD) are in both protein-coding and non-coding regions of the genome. Interpreting the effect of non-coding regulatory SNPs is challenging. We previously developed the iSNP pipeline, which determines how SNPs in transcription factor (TF) binding sites and miRNA target sites affect cellular regulatory networks and contribute to disease pathogenesis. We successfully used this method in Ulcerative Colitis, and now we developed this method further to study CD. Methods We applied the iSNP pipeline to identify CD-associated SNP affected genes in 1695 genotyped CD patients. The original iSNP approach can only analyse the direct interactors of SNP affected genes. Here, we implemented a heat propagation algorithm to uncover further downstream pathways, TFs and regulatory processes involved in the pathogenesis of CD. The heat was propagated from the proteins encoded by the SNP-affected proteins through a signalling network (using OmniPath) to TFs, and from them to their target genes (using DoRothEA). Results From the disease causing SNPs on the immunochip data of the patients, we identified 47 regulatory SNPs out of which 39 SNPs were in TF binding sites, 5 were affecting miRNA-target sites and 3 were affecting both. The pipeline predicted these SNPs regulate the expression of 83 proteins; 47 of them were present in the OmniPath signalling network. The network propagation analysis identified 518 proteins that were affected significantly in CD by these 47 proteins. The affected proteins were involved in antigen presentation through MHC-I, WNT signalling, C-type lectin signalling and NF-kB signalling pathways based on Gene Ontology analysis. GNAI1, RIPK2, ATG16L1, CLTC, and CARD9 were affected in 63% of the patients. TF-target gene analysis identified a downstream regulatory network containing 2067 potentially affected genes. These genes were involved in T cell activation, other immune functions, DNA repair and autophagy regulation. Genes encoding MMP9, RARA and SSB2 were affected in 70% of patients. Conclusion The affected signalling network showed the known importance of the WNT pathway, autophagy and type I antigen presentation in CD. We extended the list of affected processes with the network propagation modelling approach, and pointed out key functions that were commonly affected in the majority of patients with CD. Thus, patient-specific, various SNP sets converge on the same, few key functions, such as fibrosis and tissue remodelling (through MMP9 or RARA), and DNA repair (through SSB2). Overall, the new network propagation model in the iSNP pipeline pointed out genes affected in the majority of CD patients, highlighting their potential role in CD pathogenesis.