Genomic prediction models for traits differing in heritability for soybean, rice, and maize

Authors: KALER, AVJINDER - University Of Arkansas; PURCELL, LARRY - University Of Arkansas; BEISSINGER, TIMOTHY - Goettingen University; Gillman, Jason

Submitted to: BMC Plant Biology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/17/2022
Publication Date: 2/26/2022
Citation: Kaler, A.S., Purcell, L., Beissinger, T., Gillman, J.D. 2022. Genomic prediction models for traits differing in heritability for soybean, rice, and maize. Biomed Central (BMC) Plant Biology. 22. Article 87. https://doi.org/10.1186/s12870-022-03479-y.
DOI: https://doi.org/10.1186/s12870-022-03479-y

Interpretive Summary: Traditionally, plant breeders have made decisions on which lines to advance based on collection of phenotypic field data. This process is inherently expensive and high requirement for labor. In recent years, use of large genetic marker data sets, in combination with a smaller set of phenotypic data collection, permits development of mathematical models that can efficiently predict field performance. This is practice is termed Genomic selection, and it has been conclusively demonstrated to be superior to use of field phenotypic data alone. Application of this technology has the potential to dramatically increase the rate of genetic gains and accelerate essential breeding to meet the needs of farmers and the general public. Despite these known benefits, optimization of genomic selection methods is still needed. We leveraged previously developed genetic and phenotypic data for three different species that represent the range of genetic diversity potential and pollination biology in crop species: soybean corn and rice. We then tested a range of different genomic selection methods with these three datasets and using a range of different parameters. We conclude that overall genomic prediction accuracy can be significantly improved if users utilize one particular modeling method (Bayes B) with only a subset of significant markers using an appropriate training population. Our results were consistent across all three crop species and have the potential for greater accuracy and efficiency in crop breeding using genomic selection.

Technical Abstract: Background Genomic selection is a powerful tool in plant breeding. By building a prediction model using a training set with markers and phenotypes, genomic estimated breeding values (GEBVs) can be used as predictions of breeding values in a target set with only genotype data. There is, however, limited information on how prediction accuracy of genomic prediction can be optimized. The objective of this study was to evaluate the performance of 11 genomic prediction models across species in terms of prediction accuracy for two traits with different heritabilities using several subsets of markers and training population proportions. Species studied were maize (Zea mays, L.), soybean (Glycine max, L.), and rice (Oryza sativa, L.), which vary in linkage disequilibrium (LD) decay rates and have contrasting genetic architectures. Results Correlations between observed and predicted GEBVs were determined via cross validation for three training-to-testing proportions (90:10, 70:30, and 50:50). Maize, which has the shortest extent of LD, showed the highest prediction accuracy. Amongst all the models tested, Bayes B performed better than or equal to all other models for each trait in all the three crops. Traits with higher broad-sense and narrow-sense heritabilities were associated with higher prediction accuracy. When subsets of markers were selected based on LD, the accuracy was similar to that observed from the complete set of markers. However, prediction accuracies were significantly improved when using a subset of total markers that were significant at P = 0.05 or P = 0.10. As expected, exclusion of QTL-associated markers in the model reduced prediction accuracy. Prediction accuracy varied among different training population proportions. Conclusions We conclude that prediction accuracy for genomic selection can be improved by using the Bayes B model with a subset of significant markers and by selecting the training population based on narrow sense heritability.