Fine-scale recombination rates inferred using the canFam4 assembly are strongly correlated with previous maps of dog recombination

Background: Genetic maps are a key resource for the analysis of genetic variation. Prior maps in dogs were based on the canFam3.1 genome assembly, which contains thousands of sequence gaps that include functional elements. Although long-read technologies have enabled the generation of updated reference assemblies, updated maps are required to take full advantage of these new resources. Methods: We inferred fine-scale recombination maps that consider population demographic history for four populations of village dogs sequenced by the Dog10K Consortium. We compare properties among the newly generated and previously published maps and assess the effects of genetic map choice on genotype phasing accuracy. Results: The newly inferred maps are broadly consistent with previous studies of recombination in dogs. At small scales, stronger correlations are observed among maps inferred using linkage disequilibrium than with a pedigree-based map. The position of recombination hotspots showed significant overlap among inferred maps and were enriched for sequences that were gaps in a previous canine genome assembly. When using the Dog10K samples as a reference panel, a phase error rate of approximately 2% was found with the choice of genetic map having little effect. Conclusions: We describe genetic maps based on the canFam4 assembly which will be useful for future studies of genetic variation that rely on updated dog genome references.