DISSECTING COMPLEX TRAITS IN MAIZE AND BIOFUEL GRASSES BY APPLYING GENOMICS, BIOINFORMATICS, AND GENETIC RESOURCES
Location: Plant, Soil and Nutrition Research
Title: Software engineering the mixed model for genome-wide association studies on large samples
Submitted to: Briefings in Bioinformatics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: September 25, 2009
Publication Date: November 1, 2009
Citation: Zhang, Z., Buckler IV, E.S., Casstevens, T., Bradbury, P. 2009. Software engineering the mixed model for genome-wide association studies on large samples. Briefings in Bioinformatics. 10(6):664-675.
Interpretive Summary: The cost of evaluating many genetic markers in large groups of individuals has decreased rapidly in the past few years. When individuals from naturally occurring populations have been scored for genetic markers, associations can be detected between observed traits and some of those markers. That information can be used to find genes underlying variation for a trait or to develop markers to assist in breeding programs. This type of study, called association analysis, is becoming increasingly popular. Unfortunately, population structure or relatedness between individuals in populations can cause unlinked markers and traits to be associated making the results of analysis difficult to interpret. This report reviews research showing that a statistical method called the mixed model can effectively correct for population structure and greatly reduce the number of false results. In addition, this review describes software for applying mixed models to genetic data and discusses potential methods for decreasing the time required to analyze these models.
Mixed models improve the ability to detect phenotype-genotype associations in the presence of population stratification and multiple levels of relatedness in genome-wide association studies (GWAS), but for large data sets the resource consumption becomes impractical. At the same time, the sample size and number of markers used for GWAS is increasing dramatically, resulting in greater statistical power to detect those associations. The use of mixed models with increasingly large data sets depends on the availability of software for analyzing those models. While multiple software packages implement the mixed model method, no single package provides the best combination of fast computation, ability to handle large samples, flexible modeling and ease of use. Key elements of association analysis with mixed models are reviewed, including modeling phenotype-genotype associations using mixed models, population stratification, kinship and its estimation, variance component estimation, use of best linear unbiased predictors or residuals in place of raw phenotype, improving efficiency and software–user interaction. The available software packages are evaluated, and suggestions made for future software development.