Skip to main content
. Author manuscript; available in PMC: 2015 Mar 4.
Published in final edited form as: Genet Epidemiol. 2013 Jul 19;37(6):603–613. doi: 10.1002/gepi.21748

Table IV.

Distribution of the DSL rank based on the Step-1 screening test from the EDG×E and ocktail methods

Mean
Rank**
Distribution of Step-2 Testing Bins***

Model* Method 1 2 3 4 5 >=6
Base EDG×E 2.3 82% 5% 4% 2% 2% 5%
Cocktail 42.4 30% 9% 9% 8% 8% 36%
Marginal ORG
1.1 EDG×E 21.0 44% 10% 8% 6% 6% 26%
Cocktail 62.2 30% 8% 6% 7% 7% 42%
1.3 EDG×E 1.1 98% 1% 1% 0% 0% 0%
Cocktail 3.3 73% 8% 5% 3% 3% 8%
DSL-E Association
ORDSL-E=0.8 EDG×E 1,011 7% 5% 5% 5% 6% 72%
Cocktail 2,593 9% 6% 5% 6% 6% 68%
ORDSL-E=1.2 EDG×E 1.5 100% 0% 0% 0% 0% 0%
Cocktail 17.4 47% 9% 7% 6% 7% 24%
G-E Association
10 SNPs EDG×E 17.4 0% 80% 9% 3% 2% 6%
Cocktail 91.0 0% 28% 14% 10% 10% 38%
50 SNPs EDG×E 63.9 0% 0% 0% 88% 5% 7%
Cocktail 178.5 0% 0% 0% 43% 16% 41%
G-D Association
10 SNPs EDG×E 5.2 73% 12% 5% 4% 2% 4%
Cocktail 62.5 18% 17% 10% 9% 8% 38%
*

See Table 3 for a description of Models

**

Geometric mean across replicate simulations of the DSL rank based on Step 1.

Total number of SNPs = 1 million

***

Bin 1 includes the 5 top ranked SNPs based on Step 1 p-value, Bin 2 includes ranks 6 – 15, Bin 3 16–35, Bin 4 36–75, Bin 5 76–155, and the last column ranks 156 - 1 million