*2.6. Genome-Wide Association Study (GWAS)*

The GWAS was conducted in TASSEL. Several linear models were examined to achieve the best fit possible, including: general linear model (GLM) [59] using only the SNP markers as fixed e ffect; mixed linear model (MLM) [60] with kinship as a random e ffect [59]; MLM with kinship and adding principal components as covariates; and MLM with kinship and phenotypic traits (bran yield and friability) as covariates. The model fit was evaluated by visualizing QQ-plots plotted using the "qqman" package in R [61]. The model that provided the best fit was used to generate the significance values, additive e ffect scores, and the amount of phenotypic variation explained by the markers. Manhattan plots displaying the significance of the SNPs after Bonferroni correction for each of the traits were plotted using the "qqman" package in R.
