Activity Number:
|
43
- Statistical Genetics II – New Models for Complex Study Designs
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Section on Statistics in Genomics and Genetics
|
Abstract #312205
|
|
Title:
|
A Comparative Analysis of Penalized Linear Mixed Models in Structured Genetic Data
|
Author(s):
|
Anna Reisetter* and Patrick Breheny
|
Companies:
|
University of Iowa, Department of Biostatistics and University of Iowa, Department of Biostatistics
|
Keywords:
|
Genetics;
Penalized Regression;
Mixed Model;
Population Stratification
|
Abstract:
|
Many genetic studies that aim to identify genetic variants associated with complex phenotypes are subject to confounding due to unobserved factors. This poses a challenge to the detection of multivariate associations of interest and is known to induce spurious associations when left unaccounted for. Linear mixed models (LMMs) are an attractive method to correct for unobserved confounding. These methods simultaneously correct for relatedness and population structure by modeling it as a random effect with a covariance structure estimated from observed genetic data. However, population structure itself does not confound the phenotype-genotype relationship, rather differential environmental exposures among subgroups do. Population structure may or may not serve as a good proxy for such differences. Given these subtle distinctions in confounding sources, the ability of LMMs to accurately estimate fixed, sparse genetic effects has not been well studied. Considering this, we evaluate the performance of penalized LMMs in terms of MSE and false positive rates in the presence of confounding due to varying levels of environmental heterogeneity and relatedness.
|
Authors who are presenting talks have a * after their name.