Genome-wide association studies in elite varieties of German winter barley using single-marker and haplotype-based methods

Genome-wide association studies (GWAS) became a widely used method to map qualitative and quantitative traits in plants. We compared existing single-marker and haplotype-based methods for GWAS with a focus on barley. Based on German winter barley cultivars, four different single-marker and haplotype-based methods were tested for their power to detect significant associations in a large genome with a limited number of markers. We identified significant associations for yield and quality-related traits using the iSelect array with 3886 mapped single nucleotide polymorphism (SNP) markers in a structured population consisting of 109 genotypes. Genome simulations with different numbers of genotypes, marker densities and marker effects were used to compare different GWAS methods. Results of simulations revealed a higher power in detecting significant associations for haplotype- than for single-marker approaches, but showed a higher false discovery rate for SNP detection, due to lack of correction for population structure. Our simulations revealed that a population size of about 500 individuals is required to detect QTLs explaining a small trait variance (<10%).