Skip to main content

Table 2 Pairwise comparison of scoring methods surflex-dock (surf), ICM, and their consensus (maxz), at three testing fractions decided a priori. For each pair of scoring methods, differences and standard errors of estimated hit enrichment curves are shown, along with raw and Benjamini-Hochberg adjusted p-values

From: Confidence bands and hypothesis tests for hit enrichment curves

 

N tested = 3 \((r=0.001)\)

N tested = 32 \((r=0.01)\)

N tested = 321 \((r=0.1)\)

 

Diff

Std Err

Raw p

Adj p

Diff

Std Err

Raw p

Adj p

Diff

Std Err

Raw p

Adj p

EmProc

 maxz - surf

0.0000

0.0005

1.000

1.000

− 0.0118

0.0237

0.6200

0.6970

0.0588

0.0254

2.07e−02

6.21e−02

 maxz - icm

0.0118

0.0143

0.410

0.527

0.0824

0.0402

0.0407

0.0733

0.3060

0.0541

1.60e−08

1.44e−07

 surf - icm

0.0118

0.0142

0.409

0.527

0.0941

0.0429

0.0281

0.0632

0.2470

0.0626

7.91e−05

3.56e−04

McNemar

 maxz - surf

0.0000

0.0000

1.000

1.000

− 0.0118

0.0311

0.705

0.794

0.0588

0.0255

2.53e−02

7.60e−02

 maxz - icm

0.0118

0.0203

0.564

0.725

0.0824

0.0557

0.144

0.260

0.3060

0.0552

2.07e−06

1.86e−05

 surf - icm

0.0118

0.0203

0.564

0.725

0.0941

0.0614

0.131

0.260

0.2470

0.0642

3.86e−04

1.74e−03

IndJZ

 maxz - surf

0.0000

0.0138

1.000

1.000

− 0.0118

0.0497

0.8130

0.915

0.0588

0.0609

3.34e−01

5.28e−01

 maxz - icm

0.0118

0.0143

0.411

0.528

0.0824

0.0482

0.0874

0.197

0.3060

0.0668

4.74e−06

4.26e−05

 surf - icm

0.0118

0.0143

0.409

0.528

0.0941

0.0471

0.0458

0.137

0.2470

0.0693

3.63e−04

1.64e−03

CorrBinom

 maxz - surf

0.0000

0.0000

1.000

1.000

− 0.0118

0.0311

0.705

0.793

0.0588

0.0255

2.12e−02

6.35e−02

 maxz - icm

0.0118

0.0203

0.563

0.724

0.0824

0.0557

0.139

0.251

0.3060

0.0552

3.07e−08

2.76e−07

 surf - icm

0.0118

0.0203

0.563

0.724

0.0941

0.0614

0.125

0.251

0.2470

0.0642

1.20e−04

5.40e−04