Genomic Prediction across Structured Hybrid Populations and Environments in Maize.

ABSTRACT: Genomic prediction (GP) across different populations and environments should be enhanced to increase the efficiency of crop breeding. In this study, four populations were constructed and genotyped with DNA chips containing 55,000 SNPs. These populations were testcrossed to a common tester, generating four hybrid populations. Yields of the four hybrid populations were evaluated in three environments. We demonstrated by using real data that the prediction accuracies of GP across structured hybrid populations were lower than those of within-population GP. Including relatives of the validation population in the training population could increase the prediction accuracies of GP across structured hybrid populations drastically. G × E models (including main and genotype-by-environment effect) had

SUBMITTER: Li D 

PROVIDER: S-EPMC8227059 | biostudies-literature | 2021 Jun

REPOSITORIES: biostudies-literature