Currently, most PRS software only supports input of the genotyped format. Therefore, users need to generate a large intermediate file containing the best-guess genotypes and discard any information related to imputation uncertainty. To reduce the storage space requirement and to incorporate imputation uncertainty into PRS analyses, PRSice-2 implements support for the BGEN imputation format. PRSice-2 can directly process the BGEN imputed format and convert to either best-guess genotypes or dosages when calculating the PRS, without generating a large intermediate file. While PRSs based on best-guess genotypes are calculated as for genotyped input, dosage-based PRSs are calculated as (1)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}\begin{equation*} \mathrm{PRS}\ = {\mathop \sum \nolimits_i^m {\beta _i}\left( {\mathop \sum \nolimits_{j = 0}^2 {\omega _{ij}} \times j} \right)}, \end{equation*}\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${\omega _{ij}}$\end{document} is the probability of observing genotype j, where j ∈ {0,1,2}, for the i th SNP; m is the number of SNPs; and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${\beta _i}$\end{document} is the effect size of the ith SNP estimated from the relevant base genome-wide association study (GWAS) data.
