Association Studies On Yield And Quality Traits Of Maize And Rice

With fast development of whole-genome sequencing technology, more and more SNPs have been available, making it possible for us to uncover genetic architecture of complex traits by GWAS (Genome-wide association study). However, the current GWAS strategy usually ignores epistasis and gene-environment interactions due to absence of appropriate statistical methodology and a heavy computational burden. This study employed a GWAS strategy by combining the GPU-GMDR (Graphics Processing Unit-Generalized Multifactor Dimensionality Reduction) algorithm with mixed linear model approach to investigate the genetic architecture of important ear traits (ear length, ear diameter and ear row number) in maize and quality traits (gelatinization temperature, amylose content and gel consistency) in rice.For maize,305significant QTSs (quantitative trait SNPs) were detected with quite a number of additive effects, epistasis and environment interactions, and environment-specific interactions explained large proportion of phenotypic variance. Additive effects and additive-environment interactions of ear length, ear diameter and ear row number accounted for most of phenotypic variation. Wherein, environment-specific interactions contributed more, indicating all of the three ear traits were sensitive to the environments. Furthermore, we confirmed that some individual QTSs without or with small significant individual effects could affect trait expression by epistasis with other QTSs or interaction with environments.The reliability and efficiency of the model and analytical methods were verified through Monte Carlo simulations, suggesting that a population size of nearly150recombinant inbred lines had a reasonable resolution for the scenarios considered. Meanwhile, a GWAS was conducted to investigate the additive, epistasis and gene-environment associations between701,867SNPs and the three quality traits of a RIL population with138individuals derived from super-hybrid rice Xieyou9308in two environments using the above two-step strategy. Four significant SNPs, within or near the functional gene regions, were detected with additive, epistatic and gene-environment interaction effects. Our studies showed that the mixed linear model approach combining with the GPU-based GMDR algorithm is an effective strategy for implementing GWAS to uncover genetic architecture of crop complex traits.