*3.8. Evaluation of Heterozygous Rate*

The same filtering conditions with maximum missing data of 20% and MAF of 0.01%, different heterozygosity rates (starting from 5% and 10% and every 10% until 100% maximum heterozygous SNPs (Figure 7 and Supplementary Table S7) were used to filter the genotype data in our study, and different significance cutoff thresholds were used to assess the effect of the SNPs on the seed aspect ratio.

**Figure 7.** Different heterozygosity rates to filter genotype data and different significance cut-off thresholds used to assess the effect of the SNPs on seed aspect ratio.

When the genotype data filtered by a 5% to 20% maximum heterozygous rate were used for GWAS analysis, a higher specificity of the results was obtained; however, only one significance marker was evaluated at the 0.05 critical threshold for the false discovery rate (FDR) adjusted p-value. On the contrary, when a high heterozygosity rate of 30% to 100% was used for data filtering, additional significant markers were detected; however, those markers require validation.
