A Fast and Powerful Empirical Bayes Method for Genome-Wide Association Studies.

Tianpeng Chang,Huijiang Gao,Junya Li,Mang Liang,Bo Zhu,Julong Wei,Xue Gao,Bingxing An,Lingyang Xu,Xiaoqiao Wang,Yan Chen,Lupei Zhang

doi:10.3390/ani9060305

Abstract

Simple SummaryImproving statistical power and computational efficiency are always the research foci in genome-wide association studies (GWAS). In this study, we proposed a fast empirical Bayes GWAS method, which is based on the linear mixed model framework. The method is called Fast-EB-LMM in short. Results from simulation studies show that the Fast-EB-LMM has the highest power for quantitative trait nucleotides (QTNs) detection, the highest computational efficiency, and the strongest robustness, as compared with the efficient mixed model association (EMMA) and empirical Bayes (EB). Application to beef cattle population also verified the effectiveness of this method. We believe that Fast-EB-LMM is a valuable additional tool for GWAS.Linear mixed model (LMM) is an efficient method for GWAS. There are numerous forms of LMM-based GWAS methods. However, improving statistical power and computing efficiency have always been the research hotspots of the LMM-based GWAS methods. Here, we proposed a fast empirical Bayes method, which is based on linear mixed models. We call it Fast-EB-LMM in short. The novelty of this method is that it uses a modified kinship matrix accounting for individual relatedness to avoid competition between the locus of interest and its counterpart in the polygene. This property has increased statistical power. We adopted two special algorithms to ease the computational burden: Eigenvalue decomposition and Woodbury matrix identity. Simulation studies showed that Fast-EB-LMM has significantly increased statistical power of marker detection and improved computational efficiency compared with two widely used GWAS methods, EMMA and EB. Real data analyses for two carcass traits in a Chinese Simmental beef cattle population showed that the significant single-nucleotide polymorphisms (SNPs) and candidate genes identified by Fast-EB-LMM are highly consistent with results of previous studies. We therefore believe that the Fast-EB-LMM method is a reliable and efficient method for GWAS.

Highlights

Genome wide association studies (GWAS) have been carried out in humans, plants, livestock, and other species to identify statistical associations between single-nucleotide polymorphisms (SNPs)Animals 2019, 9, 305; doi:10.3390/ani9060305 www.mdpi.com/journal/animalsAnimals 2019, 9, 305 and complex traits over the past decades [1]
Since Linear mixed model (LMM) was first applied in GWAS [4], many improved LMM-based GWAS methods have been developed, e.g., the efficient mixed model association (EMMA), the genome wide efficient mixed model association (GEMMA), and the factored spectrally transformed linear mixed model (FaST-LMM)
Statistical power was used to evaluate the effectiveness of Fast-empirical Bayes method (EB)-LMM when compared with the other two

Summary

Introduction

Genome wide association studies (GWAS) have been carried out in humans, plants, livestock, and other species to identify statistical associations between single-nucleotide polymorphisms (SNPs)Animals 2019, 9, 305; doi:10.3390/ani9060305 www.mdpi.com/journal/animalsAnimals 2019, 9, 305 and complex traits over the past decades [1]. With the development of the generation and the third generation sequencing technologies, we can genotype tens of millions of SNP markers With such a high-density marker map, there are many challenges we have to face in terms of developing efficient statistical methods and computational algorithms. The FaST-LMM method improves the computational speed by choosing a selected subset of markers to capture the polygenic background effects [8]. These methods are called exact methods because the polygenic variance is re-estimated for each SNP scanned [6,9]. In contrast to the exact methods, several approximate methods have been developed by fixing the polygenic variance at the estimated value under the null model, including genomewide rapid association using mixed model and regression (GRAMMAR), efficient mixed model association expedited (EMMAX), and population parameters previously determined (P3D) [2,10,11]

Methods

Results

Discussion

Conclusion