Optimizing Trait Predictability in Hybrid Rice Using Superior Prediction Models and Selective Omic Datasets

Hybrid breeding has dramatically boosted yield and its stability in rice. Genomic prediction further benefits rice breeding by increasing selection intensity and accelerating breeding cycles. With the rapid advancement of technology, other omic data, such as metabolomic data and transcriptomic data, are readily available for predicting genetic values (or breeding values) for agronomically important traits. In the current study, we searched for the best prediction strategy for four traits (yield, 1000 grain weight, number of grains per panicle and number of tillers per plant) of hybrid rice by evaluating all possible combinations of omic datasets with different prediction methods. We conclude that, in rice, the predictions using the combination of genomic and metabolomic data generally produce better results than single-omics predictions or predictions based on other combined omic data. Inclusion of transcriptomic data does not improve predictability possibly because transcriptome does not provide more information for the trait than the sum of genome and metabolome; rather, the computational complexity is substantially increased if transcriptomic data is included in the models. Best linear unbiased prediction (BLUP) appears to be the most efficient prediction method compared to the other commonly used approaches, including LASSO, SSVS, SVM-RBF, SVP-POLY and PLS. Our study has provided a guideline for selection of hybrid rice in terms of which types of omic datasets and which method should be used to achieve higher trait predictability.