Imputation of Unordered Markers: Accuracy and Application for Genomic Selection

Sequence-based genotyping methods can generate large numbers of markers, but without a reference genome or complete linkage map these markers are unordered and typically have a large proportion of missing data. Algorithms able to impute unordered markers have not been rigorously evaluated. Using four empirical datasets, we evaluated accuracies and characterized factors affecting accuracy of four such imputation methods which were based on k-nearest neighbors, singular value decomposition, random forest regression, and expectation maximization. Each method’s impact on the genomic selection accuracy was assessed in comparison with mean imputation. Imputation accuracies were greatest when linkage disequilibrium between markers was high, genotyped individuals were closely related, and when random forest regression imputation was used. In comparison with mean imputation, all four imputation methods evaluated led to higher genomic selection accuracies when the level of missing data was high. We conclude that high levels of missing data in dense marker sets is not a major obstacle for genomic selection, even when marker order is not known.