Skip to main content
Fig. 2 | Genetics Selection Evolution

Fig. 2

From: The size and composition of haplotype reference panels impact the accuracy of imputation from low-pass sequencing in cattle

Fig. 2

Comparison of the F1 values obtained with hap.py from GATK and DeepVariant (DV) variant calls against the truth chip set for 33 samples. a Imputation improves genotype accuracy for sequence coverages lower than 20x but has little impact for sequence coverages higher than 20x. b DV has a higher F1 score for every sample than GATK for post-filter variants. The high confidence set indicates the 17 microarray genotyped samples out of the 24 samples used later as a truth set for GLIMPSE imputation. c Similar to (b) but for post-imputation variants

Back to article page