Use of genomic prediction to screen sorghum B-Lines in hybrid test crosses

Authors: KENT, MITCHELL - Texas A&M University; FONSECA, JALES - Texas A&M University; KLEIN, PATRICIA - Texas A&M University; Klein, Robert - Bob; HAYES, CHAD - Texas A&M University; ROONEY, WILLIAMS - Texas A&M University

Submitted to: The Plant Genome
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/15/2023
Publication Date: 7/16/2023
Citation: Kent, M.A., Fonseca, J.M., Klein, P.E., Klein, R.R., Hayes, C.M., Rooney, W.L. 2023. Use of genomic prediction to screen sorghum B-Lines in hybrid test crosses. The Plant Genome. Article e20369. https://doi.org/10.1002/tpg2.20369.
DOI: https://doi.org/10.1002/tpg2.20369

Interpretive Summary: The yield potential in grain sorghum hybrids has increased at a slower rate than other cereal crops including its close relative maize. While there are many reasons for this lag, increasing hybrid performance through genomic selection has the potential to accelerate the rate of genetic gain in sorghum to levels that parallel gains in hybrid maize. To address this issue, we implemented a pilot program to predict hybrid grain yield performance in sorghum using genome prediction models. This study will provide the necessary knowledge to breeders who work to exploit genomic technologies in improving grain yield of hybrid cereal crops including sorghum.

Technical Abstract: The chemical gametocide TFMSA has become an effective tool to make breeding crosses through its consistent and complete induction of temporal male sterility in experimental sorghum B-lines. Such practice increases the opportunities to identify promising B-lines because large quantities of F1 seed can be generated prior to the laborious task of B-line sterilization. Combining TFMSA technology with genomic selection could efficiently evaluate sorghum B-lines in hybrid combination to maximize the rates of genetic gain of the crop. This study used two recombinant inbred B-line populations, consisting of a total of 217 lines which were testcrossed to two R-lines to produce 434 hybrids. Population based genomic prediction models were assessed across locations using three different cross-validation schemes (CV), each with a 70% training and 30% validation sets. The validation schemes were, CV1: hybrids chosen randomly for validation; CV2: B-lines were randomly chosen, and each chosen B-line had one of the two corresponding testcross hybrids randomly chosen for the validation; and CV3: B-lines were randomly chosen, and each chosen B-line had both of the two corresponding testcross hybrids chosen for the validation. CV1 and CV2 presented the highest prediction accuracies; nonetheless, the prediction accuracies of the CV schemes were not statistically different in many environments. We determined that combining the B-line populations could improve prediction accuracies and the genomic prediction models were able to effectively rank the bottom 70% of hybrids even when genomic prediction accuracies themselves were low. Results indicate that combining genomic prediction models and TFMSA technology can effectively aid breeders in predicting B lines hybrid performance in early generations prior to the laborious task of generating A/B line pairs.