Genome-wide association studies (GWAS) are a ubiquitous tool for identifying genetic variants associated with complex traits in structured populations. During the past years, many fast GWAS algorithms based on a state-of-the-art model, namely the linear mixed model, have been published to cope with the rapidly growing data size. In this study, we provide a comprehensive overview and benchmarking analysis of commonly used GWAS algorithms. Key mathematical techniques implemented in different algorithms were summarized. Empirical data analysis with selected algorithms showed differences regarding the identification of quantitative trait loci (QTL) in several plant species. The performance of these algorithms evaluated in 10,800 simulated data sets with distinct population size, heritability and genetic architecture revealed the impact of these parameters on the power of QTL identification and false positive rate. Based on these results, a general guide on the choice of algorithms for the research community is proposed.

A comprehensive overview and benchmarking analysis of fast algorithms for genome-wide association studies