BMC Bioinformatics

Table 5 Performance comparison in simulated data with binary outcomes

From: Significance testing in ridge regression for genetic data

			Shrinkage Parameter
			Approximate test				Permutation test
			0.1	1	10	100	0.1	1	10	100
Individuals	SNPs
500	20
		TPR	0.300	0.500	0.900	0.900	0.400	0.600	0.900	0.900
		FPR	0.023	0.036	0.068	0.142	0.078	0.078	0.099	0.174
	100
		TPR	0.100	0.100	0.500	0.900	0.200	0.200	0.400	0.900
		FPR	0.024	0.037	0.046	0.087	0.050	0.052	0.058	0.115
	1000
		TPR	0.200	0.300	0.500	0.700	0.100	0.100	0.400	0.700
		FPR	0.103	0.096	0.071	0.054	0.046	0.045	0.047	0.056
	2000
		TPR	0.000	0.300	0.500	0.700	0.200	0.300	0.300	0.700
		FPR	0.008	0.056	0.081	0.063	0.052	0.049	0.048	0.055
	ALL
		TPR	0.000	0.000	0.600	0.900	-	-	-	-
		FPR	0.000	0.000	0.014	0.068	-	-	-	-
5000	20
		TPR	0.700	0.800	1.000	1.000	0.700	0.800	1.000	1.000
		FPR	0.024	0.024	0.030	0.096	0.090	0.083	0.089	0.154
	100
		TPR	0.400	0.400	0.900	1.000	0.200	0.300	0.900	1.000
		FPR	0.027	0.028	0.041	0.078	0.071	0.067	0.078	0.110
	1000
		TPR	0.200	0.300	0.600	1.000	0.100	0.200	0.600	1.000
		FPR	0.047	0.046	0.041	0.053	0.053	0.052	0.052	0.062
	2000
		TPR	0.000	0.200	0.500	1.000	0.000	0.100	0.400	1.000
		FPR	0.074	0.067	0.056	0.057	0.053	0.052	0.053	0.058

Performance comparison between a permutation test and the approximate test in simulated genotype data with case-control phenotypes. Reported are proportion of true positive and false positive results at significance threshold α = 0.05. TPR = True Positive Rate, FPR = False positive rate. Results for each simulation scenario are averaged over ten replicates.

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com