Table 5 Prediction results for biological processes on the neXtProt dataset

From: Protein function prediction as approximate semantic entailment

Method

\(F\max\)

\(S\min\)

AUPR

AUC

Naïve

0.308

32.987

0.183

0.500

MLP

0.310

32.033

0.206

0.502

MLP (ESM2)

0.336

30.044

0.305

0.682

DeepGOCNN

0.286

32.152

0.235

0.571

DeepGOZero

0.329

31.999

0.263

0.553

DeepGraphGO

0.322

31.861

0.240

0.558

TALE

0.280

32.973

0.221

0.533

SPROF-GO

0.312

31.164

0.251

0.620

DeepGO-SE

0.349

30.170

0.312

0.683

DeepGOGAT-SE

0.350

30.218

0.312

0.666

DeepGOGATMF-SE-Pred

0.339

30.653

0.293

0.694

  1. This table shows protein-centric \({F}_{\max }\), \({S}_{\min }\) and AUPR, and the class-centric average AUC. Bold values indicate best performance.
Back to article page