Skip to main content

Table 1 Function prediction accuracy comparison

From: A probabilistic framework to predict protein function from interaction data integrated with semantic knowledge

δ

1

2

3

4

5

6

functional similarity-based method

0.446

0.432

0.434

0.451

0.472

0.490

FS weighted averaging method

0.417

0.406

0.415

0.437

0.458

0.479

annotation pattern-based method

0.306

0.311

0.321

0.340

0.362

0.386

neighborhood-based chi-square method

0.294

0.302

0.319

0.343

0.370

0.398

  1. The prediction accuracy (precision) of our functional similarity-based probabilistic approach was compared to three previous methods, the FS weighted averaging method, the annotation pattern-based method and the neighborhood-based chi-square method. The methods predicted the fixed number δ of functions for each protein. Our approach outperforms the others across any δ up to 6.