Skip to main content

Table 1 Comparison between PSMVC and various protein complex detection algorithms in terms of three evaluation metrics with respect to two reference sets

From: Protein complex detection based on partially shared multi-view clustering

Methods

# complexes

# proteins

Reference sets

   

CYC2008

SGD

   

Evaluation metrics

   

FRAC

Recall

Acc

FRAC

Recall

Acc

PSMVC

1534

5508

0.712

0.706

0.814

0.607

0.598

0.699

EC-BNMF [28]

400

1936

0.577

0.558

0.763

0.530

0.497

0.681

InteHC [3]

366

2763

0.571

0.527

0.765

0.530

0.466

0.697

ClusterONE [27]

362

1394

0.337

0.353

0.559

0.333

0.333

0.512

CMC [25]

566

1391

0.442

0.468

0.523

0.388

0.420

0.475

Linkcomm [49]

1531

2640

0.399

0.492

0.549

0.399

0.455

0.516

MCODE [47]

83

952

0.166

0.139

0.435

0.109

0.094

0.388

MINE [48]

231

1247

0.337

0.312

0.526

0.295

0.275

0.497

MF-PINCoC [50]

1099

2838

0.399

0.368

0.563

0.355

0.330

0.520

PINCoC [51]

1101

4457

0.423

0.394

0.573

0.404

0.366

0.535

RANCoC [52]

1069

2797

0.436

0.406

0.596

0.410

0.379

0.542

SPICi [26]

420

2041

0.350

0.329

0.563

0.339

0.313

0.510

BT [53]

409

1286

0.509

0.463

0.749

0.508

0.461

0.678

C2S [31]

1035

4499

0.571

0.527

0.781

0.519

0.463

0.692

CACHET [29]

449

963

0.472

0.665

0.697

0.448

0.626

0.632

Hart [54]

390

1307

0.509

0.467

0.746

0.481

0.421

0.665

Pu [55]

400

1504

0.479

0.418

0.729

0.497

0.429

0.669

  1. Here “# complexes” denotes the number of complexes predicted by each algorithm, and “# proteins” denotes the number of proteins covered by the complexes predicted by each algorithm