Genome Wide Association Study for Non-normally Distributed Traits: A Case Study for Stalk Lodging in Maize

The abundance of new genomic information available has increased the ability of statistical and computational tools to study the genetic basis of agricultural traits. As such, the genome-wide association study (GWAS), in which statistical tests of association are conducted between genome-wide marker sets and traits of interest, are one of the most predominant analyses used to dissect the genetic architecture of traits. One limitation of the statistical models typically employed in a plant GWAS is the assumption that the error terms follow a normal distribution. This project uses a novel application of statistical approaches to rigorously quantify the genomic underpinnings of a non-normally distributed trait of agronomic importance, namely stalk lodging in maize. We developed an analytical pipeline that will enable us to focus on a subset of markers in which to apply a statistically rigorous but computationally intensive model that accounts for putatively false positive signals correlated with population structure and familial relatedness. Through this methodology peak associated SNPs were identified. Future research for this project will include a simulation study to test how well this approach can identify QTL associated with binomial traits.