Significant SNP’s genotypic effects on the Tamar stage fruit color
We conducted further analysis of the SNPs identified here as associated with fruit color (Table 2). Visual inspections, shows that allelic variation in these SNPs are indeed associated with light (high R/B value) or dark brown color (low R/B value) fruit (Figure 3a and Supplementary Figure 9). In our results, the R2R3 transcription factor gene (ortholog of the oil palm VIRESCENS gene -VIR gene), previously reported to regulate red or yellow fruit color at khalal stage (fresh fruit) (K. M. Hazzouri et al., 2015; Khaled M. Hazzouri et al., 2019), is present 16 kb away from the SNP LG4s65268794 on LG4. As validation, we separated our samples by the VIR genotype and compared them to their fruit color. Our results confirm that samples were classified into light brown, dark brown and mixed color of light & dark fruit based on the VIR genotype (Supplementary Figure 10). Out of 188 samples, 27 fruits sample are homozygous wild type (VIR+/VIR+) and show dark brown color fruit, while 61 are homozygous for the transposon insertion (VIRIM/VIRIM ) and show light brown color fruit. 100 have VIR+/VIRIMgenotypes that show the mixed color of the fruit (light & dark brown) in which VIRIM act as dominant or semi-dominant by interfering with expression of the wild type VIR+ allele. We also identified start codon mutation (ATG to ATA) of R2R3 transcription factor gene (called as VIRsaf) in homozygous wild-type fruit, previously reported in Hazzouri et al’s study (2019). VIR+ is haplosufficient in VIR+/VIRsaf genotype and it produces dark brown color fruit like homozygous wild type (Khaled M. Hazzouri et al., 2019).
For further understanding of color variation of dry fruit beyond the R2R3 transcription factor, we analyzed fruit color phenotype by the genotype of significant SNPs from our GWAS (Table 2) while excluding the LG4s65268794 SNP that corresponds to the genotype state of the VIR gene. The analysis was conducted in all 188 samples and then in groups separated by VIR genotype (Figure 3a, Supplementary Figure 11). Genotypes of SNP LG4s19036701 could not distinguish the fruit color in the VIRIM sample group, likely because of linkage on the same chromosome as the VIR gene, and so was excluded from this specific analysis. The results from analyses of the combined 188 samples showed significant associations (Wilcoxon test) with the color phenotype as expected (Figure 3a). To confirm the genotypic effect of these SNPs within the genetic background of the VIR genotype sample group, we performed the phenotype and genotype analysis in each group separately. We observed that even within the various VIR genotype groups that are key to red and yellow fresh fruit color, the SNPs identified here often further distinguish fruit color significantly (Figure 3b & 3c) suggesting these regions may play a role in dry fruit color or a more fine-grain role in fresh fruit color that was previously undetected. The result from the homozygous VIR wild-type fruit group, excluding the VIRsaf/VIRsaf genotype samples, (total 25 sample) shows that fruit’s darkness decreases (R/B value increases) when the sample is heterozygous for SNP LG3s906369 while darkness increases (R/B value decreases) when the sample is homozygous for the reference allele (Figure 3b). SNP LG5s14425448, LG10s12886617 and MU008982.1s1030911 have only homozygous REF alleles in the group, so did not the numbers required to show any fruit color variation. The VIRIM sample group shows that the lightness of the fruit color increases (R/B value increases) when the sample is homozygous ALT or HET allele of SNP LG3s906369, LG5s14425448, LG10s12886617 and MU008982.1s1030911 (Figure 3c). Results from both groups agree with the 188-sample group result; fruit color can be distinguished based on the genotypic variation of the SNPs. Manual fruit color analysis by the genotypes of SNP LG3s906369 on VIRIM sample group agrees with this result (Figure 4). Sample group details and genotype information of SNPs are mentioned in the Supplementary file 2.