RICOPILI automates and integrates standard GWAS analysis methods, allowing for automated cluster submission and parallelization. The pipeline unifies standard software for its functions and implements best data analysis practices, provides sensible default settings while permitting the user to flexibly customize filters, thresholds and job resources as required. The optimization of cluster resources allows computations and visualizations to be completed quickly without significant user intervention. Written predominantly in Perl and R, the pipeline is organized according to analysis modules. Each module runs in its entirety via a single command line. The main module functions include: Pre-imputation technical QC;Principal components analysis (PCA) and relatedness estimation;Genome-wide imputation of genotype probabilities and generation of best guess genotypes in PLINK format (Purcell et al., 2007);Downstream analyses, including GWAS, meta-analysis and polygenic risk scoring;Harmonizing large imputation reference panels (such as 1000 Genomes and the Haplotype Reference Consortium) to fit the architecture of RICOPILI.
