Skip to main content
. Author manuscript; available in PMC: 2011 Apr 1.
Published in final edited form as: Genet Epidemiol. 2010 Apr;34(3):258–265. doi: 10.1002/gepi.20457

TABLE II.

Imputation yield using the fixed error rate approach

Software Phase II Reference Panela Threshold b Yield
BEAGLE CEU 0.967 575
YRI 0.824 1,976
Unweighted CEU+YRI 0.832 3,847
MACH CEU 0.999 3,121
YRI 0.688 13,927
Unweighted CEU+YRI 0.946 10,341
Weighted CEU+YRI 0.939 9,025
PLINK CEU NA 0
YRI NA 0
Unweighted CEU+YRI NA 0

Error rates were fixed at 10% for these analyses.

a

The phase II CEU and YRI reference panels both consist of 60 founders. The unweighted CEU+YRI panel consists of all 120 CEU and YRI founders in a single reference panel. A weighted CEU+YRI reference panel consisted of 48 randomly chosen YRI founders and 12 randomly chosen CEU founders consistent with observed admixture proportions of ~80% and ~20%, respectively. Results shown for the weighted panel approach are averages of 30 randomly generated panels.

b

Threshold indicates the minimum posterior probability for imputing discrete genotype calls to achieve concordance between imputed and observed genotypes of 0.90, equivalent to a fixed 10% error rate. NA indicates that the program could not achieve concordance of 0.90 using the reference panel.