Table 3.

Comparison of predictive correlations (Pearson’s) among ADD, POE, and ADD-POE models

Data	$Cor{r}_{\mathbf{y},\widehat{\mathbf{y}}}^{(\mathrm{ADD}\text{-}\mathrm{POE})}-Cor{r}_{\mathbf{y},\widehat{\mathbf{y}}}^{(\mathrm{ADD})}$	$Cor{r}_{\mathbf{y},\widehat{\mathbf{y}}}^{(\mathrm{ADD}\text{-}\mathrm{POE})}-Cor{r}_{\mathbf{y},\widehat{\mathbf{y}}}^{(\mathrm{POE})}$
Mouse	0.000	0.005
${\text{Simulated}}^{\mathrm{a}}$	0.001	−0.080

ADD additive model, POE parent-of-origin model, ADD-POE parent-of-origin model where two substitution effects are modeled only to markers with significant signals on parent-of-origin effects

${}^{\mathrm{a}}$ Takes $s=0.6$, $\mathit{\rho }=0$ as a benchmarking scenario