Table 1 Performance of the different imputation methods on the apple dataset
MethodGenotype ErrorAllele ErrorRun Time, sec
Mode23.0%12.4%a
kNNib20.6%10.8%18
MF9.9%5.1%40,107
fastPHASE7.7%3.9%52,399
Beagle7.6%3.9%424
LD-kNNic7.4%3.9%104
  • kNNi, k-nearest neighbors imputation; LD-kNNi, linkage disequilibrium k-nearest neighbors imputation.

  • a Run time was under a second.

  • b Using a fixed value of k = 8.

  • c Using fixed values of k = 5 and l = 20.