GWAS analysis of grain color and tannin content in Chinese sorghum based on resequencing technology

Abstract

Sorghum is mainly used for brewing distilled liquors in China. Since grain tannins play an important role in liquor brewing, accurately understanding the relationship between grain color and tannin content can provide basis for selection standards of tannin sorghum. We resequenced a panel of 242 Chinese sorghum accessions and performed population structure and GWAS analysis to identify QTLs affecting pericarp color, testa pigment, and tannin content. Dendrogram, principal component analysis (PCA), and admixture model were used to infer population structure. Two distinct genetic sub-populations were identified according to their corresponding northern and southern geographic origin. To investigate the genetic basis of natural variation in sorghum grain color, a genome-wide association study (GWAS) with 2,760,264 SNP was conducted with multiple models (Blink, FarmCPU, GLM, and MLM). The four models identified 1090, 259, 32 and 251 SNP loci for the color of exocarp, mesocarp, testa, and tannin content, respectively. These SNPs were distributed in 440 SNP peaks. Some novel quantitative trait loci for sorghum pericarp color, testa pigments and tannin content colocalized with homologous genes in the flavonoid biosynthetic pathway from other plants, including an orthologue of maize ( Zea mays ) bz2 and homologues of Arabidopsis ( Arabidopsis thaliana ) TT2 , TT12 , and TT16 . Our investigation of the variation in grain color and tannin content in Chinese sorghum germplasm may help guide future sorghum breeding for liquor brewing.

Off-the-shelf image analysis models outperform human visual assessment in identifying genes controlling seed color variation in sorghum

“…This would be consistent with the previously reported complex architecture of Y locus which has multiple copies of R2R3 MYB genes (yellowseed3, y1 and additional pseudogenes) within the same vicinity with both genes complimentary linked to grain color in sorghum (Nida et al 2019(Nida et al , 2021). An additional independent signal (Chr01:67,840,021, highest RMIP = 0.39) approximately 4 MB upstream of a signal at Chr01:72,465,237 was identified linked to blue intensity phenotype and the first principal component (Figure 4B) and is approximately 262 kb upstream of previously reported candidate gene (Sobic.001G349900) for variation in exocarp color in Chinese sorghum germplasm (Zhang et al 2023). Identification of multiple marker-trait associations on chromosome 1 in this and previous studies suggest numerous other loci on chromosome 1 in addition y1 may contribute to seed color pigmentation in sorghum.…”

“…An analysis conducted with a much larger population of 1,386 sorghum genotypes and human visual assessment of seed color identified two signals, one corresponding to the Y locus and the other which may correspond to the Z locus as the same genomic interval was associated with variation in both seed color and mesocarp thickness (Hu et al 2019). In a smaller population of approximately 250 Chinese sorghum genotypes where high-density marker data was available from whole genome resequencing, a combination of visual assessment of seed color phenotypes and biochemical quantification of tannin concentration was sufficient to identify Y and tan1 (Zhang et al 2023). Biochemical characterization of the abundance of multiple carotenoids in the seed of the lines of the sorghum association panel identified several signals including one corresponding to zeaxanthin epoxidase .…”

Predictive metabolomics of pearl millet phenotypic traits using a germplasm panel of genetic diversity

“…Besides this, specialized metabolites could also be evaluated to predict the testa thickness. For tannin-rich sorghum, for instance, varieties with thicker testa might have higher tannin content than those darker-pigmented, thinner testa (Zhang et al 2023); the levels of phenolics and antioxidant activity are associated with thick and colored pericarp, and a pigmented testa related to S gene (Dykes et al 2005).…”