Table 3 Comparison of the performance of our predictors on the ATP-17 test set across fractional order gradients of different orders.

From: Accurate prediction of protein–ATP binding sites based on a protein pretrained large language model and a fractional-order convolutional neural network

Fractional order

Acc

Sen

Spe

Pre

MCC

AUC

1.3

0.990

0.391

0.997

0.660

0.503

0.959

1.25

0.990

0.339

0.998

0.683

0.504

0.961

1.20

0.991

0.339

0.998

0.724

0.492

0.952

1.15

0.990

0.290

0.998

0.643

0.429

0.952

1.10

0.990

0.367

0.997

0.623

0.474

0.966

1.05

0.990

0.415

0.997

0.479

0.501

0.960

1.00

0.991

0.343

0.998

0.680

0.479

0.960

0.95

0.990

0.306

0.998

0.691

0.456

0.943

0.90

0.990

0.391

0.997

0.642

0.497

0.965

0.85

0.990

0.294

0.998

0.695

0.449

0.957

0.80

0.990

0.302

0.998

0.682

0.450

0.956

0.75

0.990

0.355

0.998

0.682

0.488

0.961

0.70

0.990

0.343

0.997

0.639

0.464

0.959

0.65

0.990

0.306

0.998

0.723

0.467

0.951

Back to article page