Variable Selection and Joint Estimation of Mean and Covariance Models with an Application to eQTL Data

Table 5

Comparisons of the performance of JMCCM with SCGGM and CAPME for the experiment with highly correlated SNPs. 20 genes (m) and 20 SNPs (p) are involved. Standard errors are presented in parenthesis.

Method			SPE	SEN	Youden
JMCCM	0.239 (0.006)	0.678 (0.011)	0.910 (0.004)	0.802 (0.007)	0.712 (0.007)
SCGGM	0.402 (0.010)	0.869 (0.011)	0.825 (0.031)	0.772 (0.017)	0.597 (0.019)
CAPME	0.447 (0.007)	1.081 (0.010)	0.004 (0.001)	1.000 (0.000)	0.0004 (0.001)
JMCCM	0.115 (0.003)	0.467 (0.006)	0.893 (0.003)	0.904 (0.006)	0.797 (0.005)
SCGGM	0.286 (0.009)	0.745 (0.012)	0.786 (0.028)	0.822 (0.019)	0.609 (0.015)
CAPME	0.215 (0.003)	0.703 (0.006)	0.003 (0.001)	1.000 (0.000)	0.003 (0.001)