Population stratification (PS) and cryptic relationships have been the main sources of excessive false-positives and false-negatives in population-based association studies. Many methods have been developed to model these confounding factors and minimize their impact on the results of the genome-wide association studies (GWAS). Often time one first determines if there is a PS in the sample and then implements a method to properly account for this effect. The objective of this study was to evaluate the impact of PS on the accuracy and power of GWAS using Bayesian multiple regression methods. We conducted a GWA study in a simulated admixed population. The genome was composed of six chromosomes, each with 1000 equally spaced markers. Fifteen segregating QTL contributed to the genetic variation of a quantitative trait with heritability of 0.30. The impact of kinship and breed composition (BC) were evaluated using three methods: Single marker simple regression (SMR), Single marker mixed linear model (MLM) and Bayesian multiple regression (BMR). Each model was fitted with and without BC. Accuracy, power, false-positive rate and the positive predictive value were calculated and used for model comparisons. SMR and BMR, both without BC, were ranked as the worst and the best performing approaches, respectively. Our results showed that, while explicit modeling of kinship and BC is a must for SMR and MLM methods, BMR can disregard them and yet result in a higher power without compromising its false-positive rate. This study showed that the BMR analysis is robust to PS and to pedigree relationships among study subjects and it performs better than a single marker MLM approach. Keywords: population structure, kinship, Bayesian multiple regression, association study
Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume Electronic Poster Session - Methods and Tools - GWAS, , 965, 2018
|Download Full PDF||BibTEX Citation||Endnote Citation||Search the Proceedings|
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.