Document Type

Open Access Preprint

Abstract

On thinking quantitatively of complex diseases, there are at least three statistical strategies for association study: single SNP on single trait, gene-or region (with multiple SNPs) on single trait and on multiple traits. The third of which is the most general in dissecting the genetic mechanism underlying complex diseases underpinning multiple quantitative traits. Gene-or region association methods based on partial least square (PLS) approaches have been shown to have apparent power advantage. However, few attempts are developed for multiple quantitative phenotypes or traits underlying a condition or disease, and the performance of various PLS approaches used in association study for multiple quantitative traits had not been assessed. We, from regression perspective, exploit association between multiple SNPs and multiple phenotypes or traits through exhaustive scan statistics (sliding window) using PLS and sparse PLS (SPLS) regression. Simulations are conducted to assess the performance of the proposed scan statistics and compare them with the existed method. The proposed methods are applied to 12 regions of GWAS data from the European Prospective Investigation of Cancer (EPIC)-Norfolk study.