Image processing and genome-wide association studies in sunflower identify loci associated with seed-coat characteristics

Abstract

Sunflower seeds (technically achenes) are characterized by a wide spectrum of sizes, shapes, and colors. These traits are genetically correlated with the branching plant architecture loci, which were introgressed into restorer lines to facilitate efficient hybrid production. To break this genetic correlation between branching and seed traits, high resolution mapping of the genes that regulate seed traits is necessary. Recent progress in genomics permits acquisition of comprehensive genotyping data for a large diversity panel, yet a major constraint for exploring the genetic basis of important phenotypes across large diversity panels is the ability to screen and characterize them efficiently. Here, we implement a cost-effective image analysis pipeline to phenotype seed characteristics in a large sunflower diversity panel comprised of 287 individuals that represents most of the genetic variation in cultivated sunflower. A genome-wide association analysis was performed for seed-coat size and shape traits and significant signals were identified around genes regulating phytohormone activity. In addition, significant seed-coat color QTLs were identified and candidate genes that effect pigmentation were detected including a phytomelanin regulating gene on chromosome 17. Finally, QTLs associated with the seed-coat striped pattern were identified and phytohormone regulating candidate genes were detected. The implementation of image analysis phenotyping for GWAS allowed efficient screening of a large diversity panel and identification of valuable genetic factors effecting seed characteristics at the finest resolution to date.