A powerful application of being able to find many thousands of genetic variants in many samples using high-throughput sequencing is genome-wide association studies (GWAS) of genotype and phenotypes. GWAS is a genomic analysis set of genetic variants in different individuals or genetic lines to see whether any particular variant is associated with a trait. There are numerous techniques for doing this, but all rely on gathering data on variants in particular samples and working out each sample's genotype before cross-referencing with the phenotype in some way or other. In this recipe, we'll look at the sophisticated mixed linear model described by Yu et al in 2006 (Nature Genetics, 38:203-208). Describing the workings of the unified mixed linear model is beyond the scope of the recipe, but it is a suitable model for use in data with large sample and broad allelic diversity and is usable on plant and animal data.