Development of powerful genome-wide gene-gene and gene-environment interaction analysis

2015 Fiscal Year Final Research Report

• Project/Area Number: 25870074
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Statistical science; Human genetics
• Research Institution: Kurume University (2015); Tohoku University (2013-2014)
• Principal Investigator: Ueki Masao 久留米大学, バイオ統計センター, 准教授 (10515860)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: 遺伝的相互作用解析 / ゲノムワイド関連解析

Outline of Final Research Achievements

World-wide genome-wide association studies (GWASs) using single nucleotide polymorphisms (SNPs) have revealed that SNPs explain small portion of variability of human phenotypes, so-called the missing heritability problem. Current standard method which tests each SNP one at a time may miss contributions by gene-gene or gene-environment interactions. However, statistical methods for those hypotheses are still underdeveloped. In this project, a new statistical testing method is developed. It allows to vary the degrees of freedom for test statistics according to the observed cell frequencies. Simulation studies confirmed that a practically adequate statistical power is achieved.
In addition, for genetic prediction problem, new sparse modeling method, smooth-threshold multivariate genetic prediction (STMGP), is developed. It is very rapid even in application to large-scale data such as whole-genome sequencing data, and is shown to have higher prediction accuracy.

Free Research Field

遺伝疫学、統計遺伝学