Table 2

Power by Intensity Range.



Power by Intensity Quartile

Data
Method
Q1
Q2
Q3
Q4
Overall Power

Golden Spike
ANOVA
0.096
0.192
0.295
0.465
0.262

Combined P
0.144
0.315
0.413
0.568
0.360

Cyber-T
0.042
0.141
0.259
0.477
0.230

Median t
0.141
0.306
0.404
0.556
0.352

Moderated t
0.060
0.159
0.265
0.396
0.220

Original t
0.006
0.009
0.021
0.033
0.017

Gene Logic Tonsil
ANOVA
0.341
0.619
0.829
0.577
0.591

Combined P
0.390
0.646
0.867
0.615
0.629

Cyber-T
0.374
0.624
0.790
0.566
0.588

Median t
0.368
0.630
0.851
0.599
0.612

Moderated t
0.258
0.597
0.746
0.549
0.537

Original t
0.203
0.547
0.713
0.473
0.483

Affymetrix Latin Square
ANOVA
0.418
0.819
0.915
0.813
0.741

Combined P
0.518
0.887
0.945
0.895
0.811

Cyber-T
0.646
0.925
0.986
0.972
0.882

Median t
0.591
0.916
0.982
0.967
0.864

Moderated t
0.584
0.888
0.968
0.927
0.841

Original t
0.518
0.847
0.941
0.872
0.795

This table shows the power for each of the four intensity quartiles as well as the overall power for each of the datasets and methods. The power is calculated as the proportion of true positives that were detected while maintaining an overall false discovery rate of 0.05 or less.

Hess and Iyer BMC Genomics 2007 8:96   doi:10.1186/1471-2164-8-96