Imputation Accuracy of DogsLife Labrador Retrievers Sequenced at Varying Read Depth

Low coverage sequencing, coupled with imputation, is becoming increasingly as a method of cost-effectively producing dense genotype profiles of study populations. In many studies, the required read depth to achieve accurate imputation is modelled varying levels of read depth produced in silico. In reality, read coverage is variable across subjects' genomes. Here, we experimentally study the effects of read coverage on imputation accuracy using Labrador retrievers recruited from the DogsLife Project. Raw coverage of subjects used in this study varied between 1X, 4X, and 60X.