Skip to main content

Table 4 The performance on the MHCBench dataset in terms of AUC.

From: An effective and effecient peptide binding prediction approach for a broad set of HLA-DR molecules based on ordered weighted averaging of binding pocket profiles

Dataset

# peptides

OWA-PSSM

TEPITOPE

TEPITOPEpan

MultiRTA

NetMHCIIpan-2.0

Set 1

1017

0.768

0.766

0.764

0.713

0.765

Set 2

673

0.735

0.734

0.727

0.685

0.739

Set 3a

590

0.739

0.736

0.734

0.701

0.710

Set 3b

495

0.756

0.753

0.748

0.715

0.753

Set 4a

646

0.753

0.750

0.748

0.699

0.759

Set 4b

584

0.746

0.745

0.738

0.706

0.751

Set 5a

117

0.655

0.668

0.640

0.599

0.649

Set 5b

85

0.664

0.689

0.646

0.597

0.640

Average

 

0.727

0.730

0.718

0.677

0.721

  1. The PSSMs of TEPITOPE, TEPITOPEpan and the results of MultiRTA were obtained from their respective web servers. The prediction of NetMHCIIpan-2.0 were computed by its stand-alone software package.