Commercial poultry and pig breeding programs make use of crossbreeding schemes to exploit heterosis. The development of a reliable method for predicting heterosis will greatly improve the efficiency of these breeding schemes by reducing their dependency on time-consuming and expensive field-tests of multiple pure-line combinations. We are investigating the applicability and accuracy of using genetic markers to predict heterosis, using data on White Leghorns from ISA-Hendrix Genetics.

In our first study we used allele frequencies from 60K SNP data of 11 purebred White Leghorn layer lines (5 sire- and 6 dam-lines), and phenotypic data of 47 crosses made between those lines, some being reciprocal crosses. Results showed that heterosis can be predicted even if phenotypes of the pure lines are not available, and that the squared difference in allele frequency between the two parental pure lines is a useful predictor of heterosis, with an accuracy of about 0.5 for both egg number and egg weight. We also showed that with this accuracy, one can reduce the number of field tests by 50%.

In commercial breeding there may be possibilities to further exploit heterosis by selecting certain sires within a line that are better suited to be mated to a particular dam-line than others. So next we looked at the potential benefit of predicting heterosis at the sire level, rather than only at the line level. For this we used individual 60K SNP genotypes of 3427 sires (from 4 sire-lines) and line allele frequencies of 6 dam-lines. We derived that heterosis at the sire level is proportional to a function of the between- and within-line heterozygosity excess in the offspring of a particular sire, relative to the mean heterozygosity of the parental pure lines. Conclusions from this study were that it is possible to predict heterosis at the sire level, thus distinguishing between sires within the same pure line whose offspring will show different levels of heterosis. In our data though, variation in genome-wide predicted heterosis between sires within a line was small (about 1%); most differences were seen between lines (about 99%). However, if sire genotypes are already available, it is worthwhile to predict heterosis at the sire level rather than at the line level.


For more information, please contact


For more information, please feel free to contact the projectmanager.

Name: Esinam Amuzu