Skip to main content

Table 2 Comparison of predicted protein properties between FA-DNN and state-of-the-art algorithms on three independent test sets

From: Complementing sequence-derived features with structural information extracted from fragment libraries for protein structure prediction

Category

SPOT-1D

Spider 3

RaptorX

trRosetta

FA-DNN

MAE (Ï•)

0.312

0.337

–

–

0.303

MAE (ψ)

0.430

0.495

–

–

0.429

MAE (θ)

0.041

0.062

–

–

0.041

MAE (Ï„)

0.428

0.479

–

–

0.418

MAE (Cβ − Cβ Dist.)

–

–

8.636

8.304

7.766

  1. The MAE of real values of protein properties were evaluated. The best performance of each category is shown in bold
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us