Fig. 6: Genome-wide association study of agronomic traits and seed micronutrient levels.
Distribution of trait-associated SNPs across the genome determined by BLINK (A) and MLMM (B) models. Similarly, the distribution of trait-associated structural variants across the genome determined by BLINK (C) and CMLM models (D). The Manhattan plots display SNPs and structural variants associated with all tested traits at a significance threshold of p > 10−4. Data were derived from individual test locations (GKVK, Bangalore, and ICAR-IIMR, Hyderabad, India) as well as BLUPs calculated based on pooled data from both locations. The tested traits include days to 50% flowering (DFF), thousand seed weight (TW), grain yield per plant (GYPP), and seed concentrations of Ca, Fe, Mg, Na, P, K, and Zn. Genome-wide association analyses were performed using the BLINK and MLMM models implemented in GAPIT. Two-sided tests were applied to assess marker-trait associations and p < 0.0001 were considered significant. Heatmaps illustrate the linkage disequilibrium (LD) blocks spanning 100 kb flanking regions of Fe-associated SNP markers Chr1A_560273 (E) and Chr2A_32889668 (F). These SNP markers, identified through GWAS analysis, are located near known genes involved in the Fe metabolic pathway. LD decay was estimated based on pairwise squared correlation coefficients (r2) between SNP markers using a two-sided test.
