The post-imputation module (Supplementary Section S4 and Fig. S5) performs association analysis using imputed dosage files, meta-analysis via METAL (Willer et al., 2010), conditional analysis, polygenic risk scoring, LD score regression (Bulik-Sullivan et al., 2015) and replication analysis. Covariates (e.g. age, sex, principal components from PCA) and alternative phenotypes, including quantitative traits may be incorporated within the post-imputation module. Automated ‘clumping’ of genome-wide significant single nucleotide polymorphisms to facilitate identification of independently associated genetic loci. Publication-ready reports and visualizations such as Manhattan plots, QQ-plots, forest plots, annotated region plots and polygenic risk distributions are generated by the module as well. It is notable that genome-wide summary statistics as well as input statistics for various Manhattan and QQ-plots, as well as clumped summary statistics are automatically made available in the distribution/folder as part of the pipeline. These could then be utilized for downstream and follow-on analysis (https://docs.google.com/document/d/1jiD25BYjPAO-TLRAPkYSspiovn8wiQ29ZmZv9Pe2I2U/) (e.g. GCTA; Yang et al., 2011, Spredixcan; Barbeira et al., 2018 and FUSION; Gusev et al., 2016) for the GWAS results.
