We downloaded summary statistics for incidence29 (5956 cases and 14,927 controls) and prognosis6 (2734 cases) from the internet and analysed 7,908,787 autosomal markers present in both datasets. Our re-analysis of these results is covered by the existing ethical approvals and informed consent reported for those studies. To estimate our regression adjustment we selected a set of 29,715 LD-pruned SNPs, from a total of 1,370,154 SNPs with imputation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2 \ge 0.99$$\end{document}R2≥0.99 in both the Crohn’s disease susceptibility GWAS and our IPF survival GWAS. LD was estimated using the genotypes of our IPF survival GWAS which has similar UK ancestry to the Crohn’s disease prognosis GWAS. The pruned set is smaller than those for IPF because the Crohn’s susceptibility GWAS was imputed to the 1000 Genomes reference, which yields fewer SNPs with high imputation R2 values than the Haplotype Reference Consortium reference.
