CMStatistics 2017: Start Registration
View Submission - CMStatistics
Title: Two-phase designs for joint trait- and genotype-dependent sampling in post-GWAS regional sequencing Authors:  Radu Craiu - University of Toronto (Canada) [presenting]
Shelley Bull - University of Toronto (Canada)
Osvaldo Espin-Garcia - University of Toronto (Canada)
Abstract: In the post-GWAS era, identifying causal variants and susceptibility genes in GWAS-identified regions of association has become an important goal for researchers. Despite decreasing costs of next-generation sequencing (NGS) technologies, sequencing all subjects in large-scale studies is still prohibitive. Two-phase designs are proposed when findings from genome-wide association study (GWAS) are followed by regional sequencing that is too expensive to implement in the entire cohort. Inference is based on a semiparametric likelihood that is optimized using the EM algorithm. A GWAS-SNP serves as a surrogate covariate in inferring association between a sequence variant and a normally distributed quantitative trait (QT). Simulations are used to quantify the efficiency and power of joint QT-SNP-dependent sampling and analysis under alternative sample allocations. Joint allocation balanced on SNP genotype and extreme-QT strata yields significant power improvements compared to marginal QT- or SNP-based allocations. A sequencing study of systolic blood pressure is used to illustrate the method.