Genomic Selection for Antioxidant Production in a Panel of Sorghum bicolor and S. bicolor × S. halepense Lines

Abstract

The purpose of this work was to assess the performance of four genomic selection (GS) models (GBLUP, BRR, Bayesian LASSO and BayesB) in 4 sorghum grain antioxidant traits (phenols, flavonoids, total antioxidant capacity and condensed tannins) using whole-genome SNP markers in a novel diversity panel of Sorghum bicolor lines and landraces and S. bicolor × S. halepense recombinant inbred lines. One key breeding problem modelled was predicting the performance in the antioxidant production of new and unphenotyped sorghum genotypes (validation set). The population was weakly structured (analysis of molecular variance, AMOVA R2 = 9%), showed a significant genetic diversity and expressed antioxidant traits with a good level of variability and high correlation. The S. bicolor × S. halepense lines outperformed Sorghum bicolor populations for all the antioxidants. The four GS models implemented in this work performed comparably across traits, with accuracy ranging from 0.49 to 0.58, and are considered high enough to sustain sorghum breeding for antioxidants production and allow important genetic gains per unit of time and cost. The results presented in this work are expected to contribute to GS implementation and the genetic improvement of sorghum grain antioxidants for different purposes, including the manufacture of health-promoting and specialty foods.

Genomic Selection for Optimum Index with Dry Biomass Yield, Dry Mass Fraction of Fresh Material, and Plant Height in Biomass Sorghum

“…A diversity panel made up of a mixture of Sorghum bicolor lines and landraces and Sorghum bicolor × Sorghum halepense advanced recombinant inbred lines was used in this work in order to set up the groundwork upon which to build future germplasm improvement and cultivar development programs. A similar panel was used previously in a genome-wide linkage disequilibrium investigation in sorghum, and in genomic prediction and selection for antioxidant production in sorghum [10,44]. In these previous studies, mixing Sorghum bicolor and Sorghum bicolor × Sorghum halepense genotypes was motivated mainly by the observed weak structure of the resulting diversity panel.…”

“…The high pairwise correlation between plant height and the aboveground dry biomass yield implied that the proportion of variance shared by these traits was mostly explained by genetic causes. A perfect correlation between plant characteristics implies that genetic effects on the traits of interest are identical, which can indicate the existence of either linkage disequilibrium, pleiotropy or causal overlap, or ascertainment bias deriving from biased sampling [10].…”

“…DNA extraction and whole-genome genotyping procedures were amply described in Habyarimana and Lopez-Cruz [10]. The molecular information used in this work consisted of genotyping-by-sequencing single nucleotide polymorphisms (GBS SNPs) produced by BGI Hong Kong Company Limited.…”

“…Genomic selection (GS) showed good results in the selection of complex quantitative traits like yields [6,7] and it has been successfully implemented in plant breeding and animal husbandry [8,9]. The main features of the GS approach is the use of algorithms that combine phenotypes and high-density marker data to predict genetic merit upon which superior unphenotyped candidates are selected [10,11]. This attribute is particularly interesting as it can reduce the costs associated with evaluating trials, shorten generation interval, and bypass the extensive field progeny testing that are otherwise required to select parental lines to be used in crossing blocks [12].…”

“…For instance, a first genomic selection study in grain sorghum was reported by Hunt et al [17] for prediction of test-cross performance in individual trials. Velazco et al [12] investigated different genomic models including pedigree information for across-environment prediction of parental breeding values in productivity and adaptability traits [10].…”