Skip to main content

Table 5 The performance of five LPI prediction methods on CV1

From: LPI-deepGBDT: a multiple-layer deep framework based on gradient boosting decision trees for lncRNA–protein interaction identification

Metric

Dataset

LPI-BLS

LPI-CatBoost

PLIPCOM

LPI-SKF

LPI-HNM

LPI-deepGBDT

Precision

Dataset 1

0.8458 ± 0.0014

0.8317 ± 0.0132

0.8428 ± 0.0060

0.8757 ± 0.0086

0.7006 ± 0.0171

0.8457 ± 0.0046

Dataset 2

0.8547 ± 0.0031

0.8220 ± 0.0139

0.8537 ± 0.0065

0.8627 ± 0.0223

0.7009 ± 0.0169

0.8567 ± 0.0038

Dataset 3

0.7110 ± 0.0011

0.6871 ± 0.0060

0.7173 ± 0.0084

0.7298 ± 0.0153

0.7054 ± 0.0169

0.7089 ± 0.0115

Dataset 4

0.5653 ± 0.0088

0.4613 ± 0.0369

0.4894 ± 0.0508

0.6108 ± 0.0249

0.6624 ± 0.0501

0.5870 ± 0.0289

Dataset 5

0.7901 ± 0.0021

0.7713 ± 0.0040

0.7721 ± 0.0021

0.7517 ± 0.0098

0.7959 ± 0.0157

0.8018 ± 0.0189

Ave.

0.7534

0.7147

0.7351

0.7661

0.7130

0.7600

Recall

Dataset 1

0.6550 ± 0.0009

0.8331 ± 0.0140

0.9632 ± 0.0028

0.5932 ± 0.0156

0.7134 ± 0.0152

0.9456 ± 0.0070

Dataset 2

0.6738 ± 0.0013

0.8399 ± 0.0201

0.9628 ± 0.0043

0.5212 ± 0.0107

0.6893 ± 0.0146

0.9495 ± 0.0063

Dataset 3

0.6270 ± 0.0006

0.6154 ± 0.0241

0.7618 ± 0.0141

0.6226 ± 0.0058

0.6930 ± 0.0113

0.7649 ± 0.0249

Dataset 4

0.5328 ± 0.0074

0.3539 ± 0.0700

0.3190 ± 0.0668

0.6056 ± 0.0280

0.6342 ± 0.0396

0.3613 ± 0.0453

Dataset 5

0.7063 ± 0.0038

0.7921 ± 0.0135

0.8569 ± 0.0037

0.6727 ± 0.0037

0.6682 ± 0.0077

0.8425 ± 0.0261

Ave.

0.6390

0.6869

0.7727

0.6030

0.6796

0.7728

Accuracy

Dataset 1

0.7512 ± 0.0005

0.8310 ± 0.0071

0.8917 ± 0.0039

0.7254 ± 0.0032

0.6571 ± 0.0112

0.8964 ± 0.0032

Dataset 2

0.7620 ± 0.0018

0.8258 ± 0.0064

0.8987 ± 0.0034

0.7065 ± 0.0081

0.6474 ± 0.0088

0.8952 ± 0.0024

Dataset 3

0.6605 ± 0.0012

0.6677 ± 0.0091

0.7298 ± 0.0034

0.6544 ± 0.0092

0.6585 ± 0.0097

0.7236 ± 0.0043

Dataset 4

0.5424 ± 0.0048

0.4801 ± 0.0201

0.4972 ± 0.0306

0.5727 ± 0.0196

0.6100 ± 0.0274

0.5506 ± 0.0167

Dataset 5

0.7337 ± 0.0025

0.7785 ± 0.0067

0.8018 ± 0.0018

0.6726 ± 0.0036

0.7117 ± 0.0053

0.8129 ± 0.0132

Ave.

0.6900

0.7166

0.7638

0.6663

0.6569

0.7757

F1-score

Dataset 1

0.7381 ± 0.0012

0.8314 ± 0.0067

0.8989 ± 0.0033

0.6298 ± 0.0070

0.7069 ± 0.0148

0.8927 ± 0.0031

Dataset 2

0.7533 ± 0.0020

0.8282 ± 0.0067

0.9048 ± 0.0027

0.5828 ± 0.0117

0.6949 ± 0.0140

0.9105 ± 0.0024

Dataset 3

0.6663 ± 0.0008

0.6480 ± 0.0148

0.7377 ± 0.0034

0.5950 ± 0.0086

0.6991 ± 0.0119

0.7337 ± 0.0068

Dataset 4

0.5483 ± 0.0081

0.3812 ± 0.0573

0.3783 ± 0.0597

0.5401 ± 0.0232

0.6480 ± 0.0445

0.4397 ± 0.0362

Dataset 5

0.7458 ± 0.0030

0.7812 ± 0.0080

0.8121 ± 0.0018

0.6345 ± 0.0041

0.7264 ± 0.0061

0.8165 ± 0.0134

Ave.

0.6904

0.6940

0.7464

0.5964

0.6951

0.7586

AUC

Dataset 1

0.9192 ± 0.0005

0.8860 ± 0.0048

0.9313 ± 0.0030

0.9344 ± 0.0073

0.7774 ± 0.0147

0.9346 ± 0.0040

Dataset 2

0.9301 ± 0.0017

0.8909 ± 0.0044

0.9389 ± 0.0034

0.9199 ± 0.0149

0.7677 ± 0.0133

0.9398 ± 0.0028

Dataset 3

0.7849 ± 0.0020

0.7151 ± 0.0112

0.8223 ± 0.0029

0.8117 ± 0.0159

0.7794 ± 0.0126

0.8083 ± 0.0042

Dataset 4

0.5843 ± 0.0094

0.4726 ± 0.0270

0.4891 ± 0.0326

0.6479 ± 0.0379

0.7038 ± 0.0438

0.5790 ± 0.0207

Dataset 5

0.8738 ± 0.0028

0.8498 ± 0.0064

0.8806 ± 0.0019

0.8455 ± 0.0076

0.8718 ± 0.0074

0.8988 ± 0.0126

Ave.

0.8185

0.7629

0.8124

0.8319

0.7800

0.8321

AUPR

Dataset 1

0.8851 ± 0.0022

0.8936 ± 0.0049

0.9224 ± 0.0037

0.9196 ± 0.0092

0.8260 ± 0.0180

0.8889 ± 0.0091

Dataset 2

0.8975 ± 0.0032

0.8929 ± 0.0050

0.9266 ± 0.0044

0.8787 ± 0.0260

0.8039 ± 0.0187

0.8991 ± 0.0068

Dataset 3

0.7469 ± 0.0006

0.7024 ± 0.0109

0.8060 ± 0.0044

0.7772 ± 0.0198

0.8039 ± 0.0161

0.7792 ± 0.0070

Dataset 4

0.5851 ± 0.0109

0.5074 ± 0.0254

0.4987 ± 0.0272

0.6348 ± 0.0340

0.7435 ± 0.0689

0.5965 ± 0.0176

Dataset 5

0.8579 ± 0.0036

0.8274 ± 0.0079

0.8626 ± 0.0027

0.8364 ± 0.0170

0.8601 ± 0.0118

0.8837 ± 0.0121

Ave.

0.7945

0.7647

0.8033

0.8093

0.8075

0.8095