Fig. 2From: Assessment of the performance of hidden Markov models for imputation in animal breedingGenotype imputation accuracy of a combination of pre-phasing and haploid HMM methods based on simulated data. Unless otherwise noted there were 1000 high-density (HD) SNPs per chromosome, 200 low-density (LD) SNPs per chromosome, 100 dams genotyped at high-density and complete overlap between the high-density arrays of generations 1 and 2 and those of generations 3 and 4. We varied a the number of dams genotyped at high-density, b the number of individuals in the population, c the number of SNPs in the low-density array, d the amount of overlap between the high-density array of generations 1 and 2 and those of generations 3 and 4, and e the number of high density SNPs per chromosome keeping the ratio between high- and low-density constant (15:1)Back to article page