Table 1 Estimation accuracies from simulation study as measured by the ratio Formula of the square root of MSE and the average realized kinship (Equation 14)
Additional informationEstimatorRelationship
FSHSC1C2C3IN
NoneFormulaPLINK1.644.6610.8150.76178.7830.30
FormulaMLE1.513.557.2627.0078.6612.79
FormulaGlobal Day-Williams3.126.3812.3449.55187.687.33
FormulaClassic GRM3.004.858.3230.83116.225.92
FormulaRobust GRM2.424.337.9931.64120.565.56
FormulaTwo-step GRM2.194.267.9630.84116.245.50
LD patternFormulaNaive LD-pruned GRM3.014.707.8728.20105.465.82
LD patternFormulaLDAK2.644.107.2123.7485.576.31
LD patternFormulaLD-weighted GRM1.592.604.4417.3365.604.82
Marker orderFormulaLocal Day-Williams2.042.766.5210.7924.268.31
LD pattern + marker orderFormulaLD-weighted local Day-Williams2.102.856.7011.2723.568.39
Genetic positionFormulaHMM1.773.366.0417.9064.974.84
LD pattern + genetic positionFormulaLD-weighted HMM1.572.835.1614.5552.194.52
PedigreeFormulaPedigree kinship7.8010.4917.6337.3070.1614.73
  • All estimators compared require dense SNP genotypes and reliable sources of marker allele frequencies as input. Pedigree kinship for the six relationship types are 0.25 (FS), 0.125 (HS), 0.0625 (C1), 0.0156 (C2), 0.0039 (C3), and 0.1094 (IN), respectively.