An association transcriptomics approach to the prediction of hybrid performance

Many animal and plant species exhibit increased growth rates, reach larger sizes and, in the cases of crops and farm animals, produce higher yields when bred as hybrids between genetically differing strains, a phenomenon known as hybrid vigour or heterosis. Despite the importance of heterosis, and its extensive genetic analysis, little understanding exists of its molecular basis. We aimed to determine whether characteristics of the leaf transcriptome, as an indicator of the innate functional genetic architecture of a plant line, could be used as markers to predict heterosis and the performance of hybrids, a methodology we term Association Transcriptomics. Relationships between transcript abundance of specific genes and the values of heterosis and heterosis-dependent traits were identified and mathematical models were constructed that relate gene expression characteristics in inbred lines of Arabidopsis thaliana and maize with vegetative biomass and grain yield, respectively, in corresponding hybrids. These models were used to predict, using gene expression data, the performance of additional hybrids. The success of the application in a monocot crop of a methodology developed in a dicot model species indicates that transcriptional markers may have widespread applicability in hybrid breeding.