Table 1 . Prior density of marker effects, prior variance of marker effects, and suggested formulas for choosing hyperparameter values by model.

Model		Prior variance	Solution for scale/variance parameter
$p\left({\beta }_j |\mathbf{\omega }\right)$	Hyperparameters	$\text{Var}\left({\beta }_j |\mathbf{\omega }\right)$
Bayesian ridge regression
$N\left({\beta }_j |0,{\sigma }_{\beta }^2\right)$	${\sigma }_{\beta }^2$	${\sigma }_{\beta }^2$	${\sigma }_{\beta }^2=\frac{h^2{\sigma }_p^2}{{\text{MS}}_X}$
Bayesian LASSO
$\text{DE}\left({\beta }_j|{\sigma }^2,{\lambda }^2\right)$	$\{{\sigma }^2,{\lambda }^2\}$	$2\frac{{\sigma }^2}{{\lambda }^2}$	$\lambda =\sqrt{2\frac{\left(1-h^2\right)}{h^2}{\text{MS}}_X}$
BayesA
$t\left({\beta }_j|\text{d}.\text{f}{.}_{\beta },S_{\beta }\right)$	$\left\{\text{d}.\text{f}{.}_{\beta },S_{\beta }\right\}$	$\frac{\text{d}.\text{f}{.}_{\beta }{S}_{\beta }^2}{\text{d}.\text{f}{.}_{\beta }-2}$	$S_{\beta }^2=\frac{\left(\text{d}.\text{f}{.}_{\beta }-2\right)}{\text{d}.\text{f}{.}_{\beta }}\frac{h^2{\sigma }_p^2}{{\text{MS}}_X}$
Spike–slab
$\begin{array}{l}\pi \times N\left({\beta }_j|0,\frac{{\sigma }_{\beta }^2}{\tau }\right)+\left(1-\pi \right)N\left({\beta }_j|0,{\sigma }_{\beta }^2\right)\text{,}\\ \left(\tau >1\right)\end{array}$	$\left\{\pi ,{\sigma }_{\beta }^2,\tau \right\}$	${\sigma }_{\beta }^2\times \left[1+\pi \frac{\left(1-\tau \right)}{\tau }\right]$	${\sigma }_{\beta }^2=\left[\frac{\tau }{\tau +\pi \left(1-\tau \right)}\right]\frac{h^2{\sigma }_p^2}{{\text{MS}}_X}$
BayesC
$\pi \times 1\left({\beta }_j=0\right)+\left(1-\pi \right)N\left({\beta }_j|0,{\sigma }_{\beta }^2\right)$	$\left\{\pi ,{\sigma }_{\beta }^2\right\}$	${\sigma }_{\beta }^2\times \left(1-\pi \right)$	${\sigma }_{\beta }^2=\frac{1}{\left(1-\pi \right)}\frac{h^2{\sigma }_p^2}{{\text{MS}}_X}$
BayesB
$\pi \times 1\left({\beta }_j=0\right)+\left(1-\pi \right)t\left({\beta }_j|\text{d}.\text{f}{.}_{\beta },S_{\beta }\right)$	$\left\{\pi ,\text{d}.\text{f}{.}_{\beta },S_{\beta }\right\}$	$\left(1-\pi \right)\frac{\text{d}.\text{f}{.}_{\beta }{S}_{\beta }^2}{\text{d}.\text{f}{.}_{\beta }-2}$	$S_{\beta }^2=\frac{1}{\left(1-\pi \right)}\frac{\left(\text{d}.\text{f}{.}_{\beta }-2\right)}{\text{d}.\text{f}{.}_{\beta }}\frac{h^2{\sigma }_p^2}{{\text{MS}}_X}$

${\mathit{MS}}_x=n^{-1}{\displaystyle {\sum }_{i=1}^n{\displaystyle {\sum }_{j=1}^p{\left(x_{\mathit{ij}}-{\stackrel{-}{x}}_j\right)}^2}}$where $x_{\mathit{ij}}\epsilon \left(0,1,2\right)$ represents number of copies of the allele coded as one at the j^th (j = 1,…,p) locus of the i^th (i = 1,…,n) individual, and ${\bar{x}}_j$ is the average genotype at the j^th marker.