T1 Meta-analysis of idiopathic pulmonary fibrosis genome-wide analyses identifies three novel genetic signals associated with disease susceptibility. (12th November 2019)

• Record Type: Journal Article
• Authors: Allen, RJ; Guillen-Guio, B; Oldham, JM; Ma, SF; Dressen, A; Paynton, ML; Kraven, LM; Obeidat, M; Li, X; Braybrooke, R; Fingerlin, TE; Hall, IP; Sayers, I; Tobin, MD; Maher, TM; Schwartz, DA; Yaspan, BL; Molyneaux, PL; Flores, C; Noth, I; Jenkins, RG; Wain, LV
• Is Part Of: Thorax. Volume 74(2019)Supplement 2
• Page Start: A1
• Page End: A1
• Publication Date: 2019-11-12
• Subjects: Chest -- Diseases -- Periodicals; Thorax; Chest -- Diseases; Periodicals; Periodicals; 617.54
• DOI: 10.1136/thorax-2019-BTSabstracts2019.1
• Languages: English
• ISSNs: 0040-6376

Abstract

Abstract : Introduction: Idiopathic pulmonary fibrosis (IPF) is a devastating incurable lung disease. There have been three previous genome-wide association studies (GWAS) of IPF susceptibility. By combining these studies, we were able to perform the largest, densest and most powerful GWAS of IPF to date. Methods: We used a two-stage approach. Stage 1 consisted of a genome-wide meta-analysis of IPF across the three previous studies. European cases and controls in each of the studies was imputed using the Haplotype Reference Consortium (HRC) reference panel. We ran genome-wide analyses adjusting for 10 principal components (to adjust for fine-scale ancestry) in each study separately and meta-analysed the results. Variants with p<5 × 10 −8 in the stage 1 analysis were further analysed in two independent case-control collections (stage 2) and were defined as significantly associated with IPF risk if they were significant after multiple testing adjustments. Statistical fine-mapping and functional follow-up using three eQTL databases was used to identify putative causal genes. Finally, we used a polygenic risk score approach to determine the contribution to IPF disease risk of genetic variants that have not been reported as associated with IPF risk. For that, variants that were located near known IPF risk signals were excluded from the score and the number of variants included was varied. Results: GWAS were performed on 10, 790, 934 genetic variants in 2, 668 IPF cases and 8, 591 controls (stage 1) with replication in 1, 467 IPF cases and 11, 874 controls (stage 2). We identified three novel signals associated with IPF susceptibility. These three novel signals were associated with decreased expression of DEPTOR (in lung tissue), KIF15 and MAD1L1 (in non-lung tissue). We replicated 11 of the previously reported 17 IPF risk variants. The most significant risk score was found to include over 800, 000 independent variants that were non-significant in our GWAS and explained about 2% of the phenotypic variation. Conclusion: The novel signals support the importance of mTOR signalling and suggest a possible role of spindle-assembly genes in IPF susceptibility. Risk score analyses suggest there are potentially hundreds of genetic variants associated with IPF susceptibility that have not yet been identified by GWAS.