Improving predictions of accuracy for hill sheep

SRUC research looking at accuracy values in the National Hill Sheep Breeding Evaluation has found a better way to account for the impact of genotyping on the accuracy values that are published for EBVs and Indexes.

The motivation behind this work was to disseminate the benefits of using genomic data in our estimation of accuracy for the relatives of genotyped animals that are themselves non-genotyped.

The work completed by Samir Id-Lahoucine at SRUC has shown that whilst the accuracy values for genotyped animals were unchanged, the accuracy values for non-genotyped animals that had genotyped progeny typically rose by 8-10 points. At a flock level the new approach does a better job of accounting for the availability of genomic information when calculating accuracy values.

Samuel Boon, Signet Manager states “This work doesn’t directly influence the genomic breeding values that Signet produce, but it does give us a much better idea of the reliance we can place upon them when making breeding decisions. It highlights the value in getting a ram’s progeny genotyped; it enables Signet to publish a breeding value for that sire which is both more informative and more reliable”