Hybrid Breeding of Rice via Genomic Selection

Hybrid breeding is the main tool to improve yield in many agricultural crops, especially in rice and maize. Genomic hybrid breeding is a technology that uses whole genome markers to predict future hybrids and predicted superior hybrids are then field evaluated. This will increase the opportunity of selecting true superior hybrids with minimum cost. Here we used the genomic best linear unbiased prediction (BLUP) to perform hybrid prediction using an existing rice population of 1495 hybrids. Replicated 10-fold cross-validations (CV) showed that the prediction accuracies (correlation between observed and predicted phenotypes) of 10 agronomic traits ranged from 0.35 to 0.92. Using the 1495 hybrids rice as a training sample, we predicted six agronomic traits of 210 hybrid rice derived from diallel crosses of 21 unrelated parents. The prediction accuracies were relatively high, varying from 0.34 (yield) to 0.86 (grain length). We concluded that the current population of 1495 hybrids can be used to predict hybrids from unrelated parents. Eventually, we used this training population to predict all potential hybrid rice of cytoplasm male sterile lines from 3000 rice varieties genotyped by the 3000 rice genome project. Top and bottom 200 predicted hybrid rice based on an index combining eight economic traits were identified. These hybrids are ready to be field tested. SNP genotypes of the training population and parameters estimated from this training population are available for general use in genomic hybrid prediction of all potential hybrids generated from all varieties of rice.