Imputation strategies for near real-time genomic prediction using nanopore sequencing

Abstract Background Genomic prediction describes the use of SNP genotypes to predict complex phenotypes and has been widely applied in humans and agriculture species. Genotyping-by-sequencing, a method which uses low-coverage sequence data paired with genotype imputation, is becoming increasingly popular for SNP genotyping. The development of Oxford Nanopore Technologies’ (ONT) MinION sequencer has now made genotyping-by-sequencing portable and rapid. Here we evaluate the speed and accuracy of genomic predictions using low-coverage ONT sequence data in a population of cattle using four imputation approaches. We also investigate the effect of SNP reference panel size on their performance. Results SNP array genotypes and ONT sequence data for 64 beef heifers were used to calculate genomic estimated breeding values (GEBVs) from 641k SNP for four traits. Accuracy of the GEBVs was much higher when flanking SNP from sequence data was used to help impute the 641k panel used for genomic predictions. Using the imputation package QUILT, correlations between ONT and low-density SNP array genomic breeding values were greater than 0.91 and up to 0.97 for sequencing coverages as low as 0.1x using a panel of 48 million SNP that flanked the 641k in the prediction equation. Imputation time was significantly reduced by decreasing the number of flanking sequence SNP used in imputation for all methods. Genomic breeding values calculated using QUILT also had higher correlations to high density SNP arrays than genomic breeding values from imputed-low density arrays for coverages as low as 0.5x. Conclusions Here we demonstrated accurate genomic prediction is possible with ONT sequence data from sequencing coverages as low as 0.1x, and imputation time can be as short as 10 minutes per sample.