The impact of incorrectly recorded parentage on inferred genotypes over multiple generations

Abstract

The absence of selective pressure against recessive deleterious mutations in the heterozygous state means that virtually every individual will carry several such mutations which have arisen over time. The inflation in frequency of a few of these mutations due to selective processes during domestication and breed formation have left modern domestic dog breeds with a high burden of genetic disease due to mutations at single genes. This has stimulated research into the causal mutations, and a consequential market in DNA tests, which enable breeders to distinguish heterozygotes from wild type homozygotes and determine pairings that will avoid producing diseased progeny. The genotypes of progeny of parents with known genotypes themselves may in some cases be definitively inferred. Importantly, two parents homozygous for non-disease causing alleles, will produce progeny with the same genotype, which may be assigned to the offspring (e.g. as ‘hereditary clear’) without the need for further testing. However, the veracity of assigned genotypes is dependent on the parentage being recorded without error, which is not the case in most species. Simulations presented here demonstrate that a modest rate of false paternity can result in a notable proportion of ‘hereditary clear’ assignments being false when ‘hereditary clear’ status is assigned across a number of generations (error rates exceeding 5% after 6 generations with a disease causing mutation frequency of 0.2). Erroneous assignment of ‘hereditary clear’ genotypes risks the production of puppies with the very disease for which a DNA test is available allowing avoidance. In light of these findings and to reduce the risks of producing puppies destined to be affected by such diseases, the Kennel Club has determined to limit the assignment of ‘hereditary clear’ status of registered dogs to 2 generations, with effect from January 2022.