Skip to main content

Table 7 Q8 prediction using the ensemble model on the CB513 dataset

From: Prediction of 8-state protein secondary structures by a novel deep learning architecture

Q8 Label

precision

recall

 

eCRRNN*

CRRNN

NCCNNa

DCRNNa

eCRRNNa

CRRNN

NCCNNa

DCRNNa

H

0.872

0.860

0.846

0.832

0.935

0.926

0.936

0.933

B

0.582

0.466

0.786

0.554

0.078

0.081

0.047

0.026

E

0.804

0.797

0.776

0.753

0.867

0.831

0.837

0.828

G

0.554

0.466

0.528

0.429

0.398

0.371

0.29

0.252

I

0

0

0

0

0

0

0

0

T

0.603

0.556

0.591

0.559

0.586

0.555

0.542

0.522

S

0.563

0.494

0.621

0.518

0.352

0.332

0.237

0.249

L

0.626

0.603

0.570

0.573

0.697

0.658

0.707

0.652

  1. aindicates ensemble model
  2. Boldface numbers indicate best performance