Multi-trait genomic prediction of yield-related traits in US soft wheat under variable water regimes

Authors: GUO, JIA - University Of Florida; KHAN, JAHANGIR - University Of Florida; PRADHAN, SUMIT - University Of Florida; SHAHI, DIPENDRA - University Of Florida; KHAN, NAEEM - University Of Florida; AVCI, MUHSIN - University Of Florida; MCBREEN, JORDAN - University Of Florida; HARRISON, STEPHEN - Louisiana State University; Brown-Guedira, Gina; MURPHY, J - North Carolina State University; JOHNSON, JERRY - University Of Georgia; MERGOUM, MOHAMED - University Of Georgia; MASON, RICHARD - University Of Arkansas; IBRAHIM, AMIR - Texas A&M University; SUTTON, RUSSEL - Texas A&M University; GRIFFEY, CARL - Virginia Polytechnic Institution & State University; BABAR, M - University Of Florida

Submitted to: Genes
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/20/2020
Publication Date: 10/28/2020
Citation: Guo, J., Khan, J., Pradhan, S., Shahi, D., Khan, N., Avci, M., Mcbreen, J., Harrison, S.A., Brown Guedira, G.L., Murphy, J.P., Johnson, J., Mergoum, M., Mason, R.E., Ibrahim, A.M., Sutton, R., Griffey, C., Babar, M.A. 2020. Multi-trait genomic prediction of yield-related traits in US soft wheat under variable water regimes. Genes. 11(11), 1270. https://doi.org/10.3390/genes11111270.
DOI: https://doi.org/10.3390/genes11111270

Interpretive Summary: Genomic prediction is a breeding method where DNA marker data on lines is used to predict performance of untested lines. Performance of genomic prediction on genetically correlated traits can be improved through an use of multi-trait prediction models under a multi-environment context. In this study, a panel of 237 soft facultative wheat (Triticum aestivum L.) lines was evaluated to compare single- and multi-trait models for predicting grain yield, harvest index, spike fertility, and thousand grain weight. The panel was grown in two locations and two years in Florida under drought and moderately drought stress conditions, while the genotyping was performed using 27,957 genotyping-by-sequencing (GBS) makers. Five predictive models including Multi-environment Genomic Best Linear Unbiased Predictor (MGBLUP), Bayesian Multi-trait Multi-environment (BMTME), Bayesian Multi-output Regressor Stacking (BMORS), single-trait Multi-environment Deep Learning (SMDL), and Multi-trait Multi-environment Deep Learning (MMDL) were compared. Across environments, the multi-trait statistical model (BMTME) was superior to multi-trait deep learning model for prediction accuracy in most scenarios, but the deep learning models were comparable to other models for response to selection. The multi-trait model also showed 5 to 22% more genetic gain compared to the single-trait model across environments. Overall, these results suggest that multi-trait genomic prediction can be an efficient strategy for improving economically important yield related traits in soft wheat.

Technical Abstract: Performance of genomic prediction (GP) on genetically correlated traits can be improved through an interdependence multi-trait model under a multi-environment context. In this study, a panel of 237 soft facultative wheat (Triticum aestivum L.) lines was evaluated to compare single- and multi-trait models for predicting grain yield (GY), harvest index (HI), spike fertility (SF), and thousand grain weight (TGW). The panel was phenotyped in two locations and two years in Florida under drought and moderately drought stress conditions, while the genotyping was performed using 27,957 genotyping-by-sequencing (GBS) SNP makers. Five predictive models including Multi-environment Genomic Best Linear Unbiased Predictor (MGBLUP), Bayesian Multi-trait Multi-environment (BMTME), Bayesian Multi-output Regressor Stacking (BMORS), single-trait Multi-environment Deep Learning (SMDL), and Multi-trait Multi-environment Deep Learning (MMDL) were compared. Across environments, the multi-trait statistical model (BMTME) was superior than multi-trait DL model for prediction accuracy in most scenarios, but the DL models were comparable to the statistical models for response to selection. The multi-trait model also showed 5 to 22% more genetic gain compared to the single-trait model across environment reflected by the response to selection. Overall, these results suggest that multi-trait genomic prediction can be an efficient strategy for economically important yield component related traits in soft wheat.