Figure 5.
Estimates of expected heterozygosity (He; A), inbreeding coefficient (Fst; B), Phi statistic (Φst; C), and topology accuracy (D) obtained from 100 runs of simulation on the wheat SNP genotype data set with 10–90% of total observations as random missing and as imputed by the probabilistic principal component analysis (PP) imputation for four sample size combinations of breeding lines (L = 130, 260) and SNP markers (m = 1000, 10,000).