Robustification of Linear Regression and Its Application in Genome-wide Association Studies

A total of 138 RILs (recombinant inbred lines) population was used for the association study of both rice quality traits chalkiness degree (CD) and chalkiness percentage (CP) in this study. The SNPs presented in the data were obtained after removing the low-frequency SNPs (minor allele frequency <5%) and pruning LD correlated SNPs (r2 > 0.4) using Plink. Subsequently, we have applied generalized multifactor dimensionality reduction procedure to screen potential SNPs associated with two rice grain quality traits chalkiness degree (CD) and chalkiness percentage (CP). Finally, we have obtained 690 SNPs for CD and 619 SNPs for CP. We applied our proposed, least-squares (LS), re-weighted LS (RLS) and fast-S methods to identify the important SNPs associated with CD and CP.