Skip to main content

Table 3 Five-Fold cross-validation results on four gold-standard datasets through DeepMPF

From: DeepMPF: deep learning framework for predicting drug–target interactions based on multi-modal representation with meta-path semantic analysis

Dataset

Fold

Acc

Prec

Sen

F1

MCC

Enzyme

1

0.9111

0.9242

0.8957

0.9097

0.8226

2

0.9111

0.9397

0.8786

0.9081

0.8240

3

0.8973

0.9218

0.8682

0.8942

0.7959

4

0.9034

0.9083

0.8974

0.9028

0.8069

5

0.9058

0.9217

0.8870

0.9040

0.8122

Average

0.9057 ± 0.0058

0.9231 ± 0.0112

0.8854 ± 0.0122

0.9038 ± 0.0061

0.8123 ± 0.0116

GPCR

1

0.7756

0.8571

0.6614

0.7466

0.5661

2

0.8031

0.8667

0.7165

0.7845

0.6156

3

0.8228

0.8596

0.7717

0.8133

0.6491

4

0.7756

0.8365

0.6850

0.7532

0.5604

5

0.8031

0.8291

0.7638

0.7951

0.6082

Average

0.7960 ± 0.0203

0.8498 ± 0.0161

0.7197 ± 0.0481

0.7785 ± 0.0282

0.5999 ± 0.0369

Ion channel

1

0.9375

0.9360

0.9392

0.9376

0.8750

2

0.9407

0.9610

0.9186

0.9393

0.8822

3

0.9169

0.9271

0.9051

0.9160

0.8341

4

0.9254

0.9435

0.9051

0.9239

0.8516

5

0.9322

0.9264

0.9390

0.9327

0.8645

Average

0.9305 ± 0.0096

0.9388 ± 0.0143

0.9214 ± 0.0171

0.9299 ± 0.0098

0.8615 ± 0.0192

Nuclear receptor

1

0.8333

0.9286

0.7222

0.8125

0.6838

2

0.7222

0.7857

0.6111

0.6875

0.4558

3

0.6667

0.7500

0.5000

0.6000

0.3536

4

0.6944

0.7059

0.6667

0.6857

0.3895

5

0.8333

0.8750

0.7778

0.8235

0.6708

Average

0.7500 ± 0.0786

0.8090 ± 0.0913

0.6556 ± 0.1069

0.7219 ± 0.0947

0.5107 ± 0.1565