Polygenic risk scores (PRSs) have become an increasingly popular approach for demonstrating polygenic influences on complex traits and for establishing common polygenic signals between different traits. PRSs are typically constructed using pruning and thresholding (P+T), but the best choice of parameters is uncertain; thus multiple settings are used and the best is chosen. This optimization can lead to inflated type I error. To correct this, permutation procedures can be used but they can be computationally intensive. Alternatively, a single parameter setting can be chosen a priori for the PRS, but choosing suboptimal settings result in loss of power. We propose computing PRSs under a range of parameter settings, performing principal component analysis (PCA) on the resulting set of PRSs, and using the first PRS-PC in association tests. The first PC reweights the variants included in the PRS with new weights to achieve maximum variation over all PRS settings used. Using simulations, we compare the performance of the proposed PRS-PCA approach with a permutation test and a priori selection of p-value threshold. We then apply the approach to the Mayo Clinic Bipolar Disorder Biobank study to test for PRS association with psychosis using a variety of PRSs constructed from summary statistics from the largest studies of psychiatric disorders and related traits. The PRS-PCA approach is simple to implement, outperforms the other strategies in most scenarios, and provides an unbiased estimate of prediction performance. We therefore recommend it to be used PRS association studies where multiple phenotypes and/or PRSs are being investigated.

A principal component approach to improve association testing with polygenic risk scores