Skip to main content

Table 1 The mean square errors (MSE) of estimated gene expression levels (up:down = 9:1) for simulated cDNA microarray data with the R-I plots similar to Figure 2.

From: A robust two-way semi-linear model for normalization of cDNA microarray data

Percentage of DEG

Descriptive Statistics

 

Method

Mean

Minimum

25% Quantile

Median

75% Quantile

Maximum

1%

OLS

0.0837

0.0243

0.0474

0.0614

0.1074

1.7586

 

Huber

0.0510

0.0106

0.0235

0.0312

0.0703

1.1750

 

Tukey

0.0481

0.0101

0.0215

0.0291

0.0675

0.8684

 

LOWESS

0.0849

0.0197

0.0485

0.0642

0.1085

1.7488

5%

OLS

0.0984

0.0197

0.0487

0.0648

0.1128

2.1413

 

Huber

0.0605

0.0117

0.0246

0.0331

0.0740

1.5012

 

Tukey

0.0556

0.0110

0.0226

0.0305

0.0712

1.2406

 

LOWESS

0.0990

0.0234

0.0493

0.0677

0.1145

2.1275

10%

OLS

0.1198

0.0259

0.0519

0.0695

0.1244

1.7445

 

Huber

0.0749

0.0118

0.0266

0.0371

0.0830

1.1705

 

Tukey

0.0673

0.0111

0.0241

0.0340

0.0795

1.0206

 

LOWESS

0.1196

0.0232

0.0514

0.0722

0.1264

1.7935

20%

OLS

0.1545

0.0239

0.0601

0.0855

0.1482

2.2914

 

Huber

0.0983

0.0121

0.0322

0.0497

0.0994

1.4550

 

Tukey

0.0854

0.0112

0.0285

0.0451

0.0932

1.1777

 

LOWESS

0.1530

0.0244

0.0602

0.0900

0.1612

2.1958

40%

OLS

0.2099

0.0287

0.0835

0.1293

0.2086

2.3221

 

Huber

0.1365

0.0155

0.0465

0.0827

0.1428

1.6918

 

Tukey

0.1164

0.0135

0.0395

0.0735

0.1296

1.4402

 

LOWESS

0.2220

0.0345

0.1153

0.1665

0.2491

2.8279

  1. DEG: differentially expressed genes. OLS: the TW-SLM using the ordinary least squares.
  2. Huber: the robust TW-SLM using Huber's weight function. Tukey: the robust TW-SLM using Tukey's weight function.