Activity Number:
|
375
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 2, 2016 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistics in Epidemiology
|
Abstract #321053
|
|
Title:
|
A Stochastic Search Algorithm to Find Multi-SNP Effects Using Nuclear Families
|
Author(s):
|
David M. Umbach* and Min Shi and Alison Wise and Clare Weinberg and Leping Li
|
Companies:
|
National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences
|
Keywords:
|
evolutionary algorithm ;
genetic association study ;
epistasis ;
oral cleft
|
Abstract:
|
Given that biologic systems often involve failure-resistant redundancy, phenotypes such as birth defects may largely occur only through the joint effects of exposures and several genetic variants. Such joint effects tend to produce a weak signal in typical analyses of genome-wide association studies (GWAS), where associations are assessed one single-nucleotide polymorphism (SNP) at a time. We describe an approach that uses case-parent triads and applies the evolutionary algorithm to stochastically search the large sample space that includes all sets of size S of potentially-interacting SNPs. We assess the performance of our algorithm using realistically simulated data from the dbGaP GWAS data on families with the birth defect oral cleft. We simulate specific multi-SNP causal effects and then try to recover those causative complexes de novo. Our method shows promising results in scenarios that involve two sets of four interacting SNPs, against a background of random cases. Once the method is fully developed and its performance described, we return to the original cleft study data to search a large set of candidate SNPs for epistatic effects related to risk of oral cleft.
|
Authors who are presenting talks have a * after their name.