Prediction of heterosis using genome-wide SNP-marker data: application to egg production traits in white Leghorn crosses

Prediction of heterosis has a long history with mixed success, partly due to low numbersof genetic markers and/or small data sets. We investigated the prediction of heterosisfor egg number, egg weight and survival days in domestic white Leghorns, using∼400 000 individuals from 47 crosses and allele frequencies on∼53 000 genome-wide single nucleotide polymorphisms (SNPs). When heterosis isdue to dominance, and dominance effects are independent of allele frequencies, heterosisis proportional to the squared difference in allele frequency (SDAF) between parental purelines (not necessarily homozygous). Under these assumptions, a linear model includingregression on SDAF partitions crossbred phenotypes into pure-line values and heterosis,even without pure-line phenotypes. We therefore used models where phenotypes of crossbredswere regressed on the SDAF between parental lines. Accuracy of prediction was determinedusing leave-one-out cross-validation. SDAF predicted heterosis for egg number and weightwith an accuracy of ∼0.5, but did not predict heterosis for survival days. Heterosispredictions allowed preselection of pure lines before field-testing, saving∼50% of field-testing cost with only 4% loss in heterosis. Accuraciesfrom cross-validation were lower than from the model-fit, suggesting that accuraciespreviously reported in literature are overestimated. Cross-validation also indicated thatdominance cannot fully explain heterosis. Nevertheless, the dominance model hadconsiderable accuracy, clearly greater than that of a general/specific combiningability model. This work also showed that heterosis can be modelled even when pure-linephenotypes are unavailable. We concluded that SDAF is a useful predictor of heterosis incommercial layer breeding.