Study of Regions of Homozygosity (ROH) Patterns to Evaluate the Use of Dogs’ Genome in Human Drug Development

Abstract

Animals are used as preclinical models for human diseases in drug development. Dogs, especially, are used in preclinical research to support clinical safety evaluations during drug development. Comparisons of patterns of regions of homozygosity (ROH) and phenotypes between dogs and humans are not well known. We conducted a genome-wide homozygosity analysis (GWHA) on the human and dog genomes. We calculated ROH patterns across distinct human cohorts, including the Amish, the 1000 genomes, Vanda 1 k genomes, and the Alzheimer’s cohort. The Amish provided a large cohort of extended kinships, allowing for in-depth family-oriented analyses. The remaining human cohorts served as statistical references. We then calculated ROH across different dog breeds, with emphasis on the beagle—the preferred breed used in drug development. Out of five studied human cohorts, we reported the highest mean ROH in the Amish population. We calculated the extent of the genome covered by ROH (FROH) (human 3.2 Gb, dog 2.5 Gb). Overall, FROH differed significantly between the Amish and the 1000 genomes, and between the human and the beagle genomes. The mean FROHper 1 Mb was ~ 16 kb for Amish, ~ 0.6 kb for Vanda 1 k, and ~ 128 kb for beagles. This result demonstrated the highest degree of inbreeding in beagles, far above that of the Amish, one of the most inbred human populations. ROH can contribute to inbreeding depression if it contains deleterious variants that are fully or partially recessive. The differences in ROH characteristics between human and dog genomes question the applicability of dog models in preclinical research, especially when the goal is to gauge the subtle effects on the organism’s physiology produced by candidate therapeutic agents. Importantly, there are huge differences in a subset of ADME genes, specifically the cytochrome P450 family (CYPs), which constitute major enzymes involved in drug metabolism. We should use caution when generalizing from dog to human, even if human and beagle are relatively close species phylogenetically. Doi: 10.28991/HEF-SP2022-01-02 Full Text: PDF