Online Program Home
My Program

Abstract Details

Activity Number: 535 - Contributed Poster Presentations: Section on Statistics in Genomics and Genetics
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #330241
Title: Inferences in High-Dimensional Misspecified Mixed Model Analysis for GWAS
Author(s): Cecilia Dao* and Jiming Jiang and Debashis Paul and Hongyu Zhao
Companies: Yale Univ and University of California, Davis and UC Davis and Yale
Keywords: asymptotic variance; misspecified LMM; REML; heritability confidence interval
Abstract:

We explore inferences for the restricted maximum likelihood (REML) estimator of the variance components under a misspecified linear mixed model (MLMM) often used in genome-wide association studies (GWAS). Through some asymptotic analysis, we propose suitable variance estimators and large sample confidence intervals for the REML estimators of the variance components and heritability of a trait under a MLMM. Extensive simulation studies show that the proposed variance estimators and confidence intervals perform well in large samples based on the percentage of relative bias and empirical coverage probabilities. We compare our method with the standard way to estimate standard errors of REML variance components and confidence interval construction along with some recently proposed methods, and show that our method is more accurate.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program