Skip to main content
ARS Home » Pacific West Area » Pullman, Washington » Plant Germplasm Introduction and Testing Research » Research » Publications at this Location » Publication #383977

Research Project: Management of Priority Legume, Oilseed, Vegetable, Forage Grass, Sugar, Ornamental, and Medicinal Plant Genetic Resources and Associated Information

Location: Plant Germplasm Introduction and Testing Research

Title: Harnessing genetic diversity in the USDA pea (Pisum sativum L.) germplasm collection through genomic prediction

Author
item AL BARI, MD. ABDULLAH - North Dakota State University
item ZHENG, PING - Washington State University
item WORRAL, HANNAH - North Dakota State University
item SZWIEC, STEPHEN - North Dakota State University
item MA, YU - Washington State University
item MAIN, DORRIE - Washington State University
item Coyne, Clarice - Clare
item McGee, Rebecca
item BANDILLO, NONOY - North Dakota State University

Submitted to: Frontiers in Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/26/2021
Publication Date: 12/24/2021
Citation: Al Bari, M., Zheng, P., Worral, H., Szwiec, S., Ma, Y., Main, D., Coyne, C.J., McGee, R.J., Bandillo, N. 2021. Harnessing genetic diversity in the USDA pea (Pisum sativum L.) germplasm collection through genomic prediction. Frontiers in Genetics. 12:707754. https://doi.org/10.3389/fgene.2021.707754.
DOI: https://doi.org/10.3389/fgene.2021.707754

Interpretive Summary: Trait evaluation and efficient utilization of germplasm collections can be time-intensive, laborious, and expensive. Genomic selection (GS), a decade-old technology to date, takes advantage of high-density genomic data and holds a promise of potentially increasing the rate of progress in plant breeding. As genotyping costs have significantly declined relative to phenotyping (trait evaluation) costs currently, GS has become an attractive option as a selection decision tool to evaluate accessions in extensive germplasm collections. A genomic prediction approach could use only genomic data to predict each accession's breeding value held in the collection. The predicted values would significantly increase the value of accession in germplasm collections by giving breeders a means to identify those favorable accessions meriting their attention among the thousands of available accessions in germplasm collections. In this study, we applied and evaluated genomic selection's potential to a set of 482 pea accessions that were genotyped with 30,600 SNP markers and phenotyped for seed yield and yield-related components to enhance selection of accessions from the USDA Pea Germplasm Collection. Desirable breeding values with higher reliability were found for days to flowering and plant height and can be used to identify and screen favorable germplasm accessions. Predictions for yield may be improved by expanding the training set of accessions used in the modeling.

Technical Abstract: Phenotypic evaluation and efficient utilization of germplasm collections can be time-intensive, laborious, and expensive. However, with the plummeting costs of next-generation sequencing and the addition of genomic selection to the plant breeder's toolbox, we now can more efficiently tap the genetic diversity from germplasm collections. In this study, we applied and evaluated genomic selection's potential to a set of 482 pea accessions that were genotyped with 30,600 SNP markers and phenotyped for seed yield and yield-related components to enhance selection of accessions from the USDA Pea Germplasm Collection. Genomic prediction models and several factors affecting predictive ability were evaluated in a series of cross-validation schemes across complex traits. Different genomic prediction models gave similar results, with predictive ability across traits ranging from 0.12 to 0.57, with no model working best across all traits. Increasing the training population size did not significantly improve the predictive ability for seed yield. However, an increasing trend in predictive ability was observed with an increasing number of SNPs. Accounting for population structure effects did not significantly boost predictive ability, but we observed a slight improvement in seed yield. By applying the best genomic prediction model from this study, we then examined the distribution of non-phenotyped accessions and the reliability of genomic estimated breeding values (GEBV) in USDA Pea accessions that were genotyped but not phenotyped. The distribution of GEBV suggested that none of those non-phenotyped accessions were expected to perform outside the range of those phenotyped accessions. Desirable breeding values with higher reliability can be used to identify and screen favorable germplasm accessions. Expanding the training set and incorporating additional orthogonal information into the genomic prediction framework could enhance prediction accuracy.