Skip to main content

Table 2 Comparison of the proposed method with pre-trained protein language model for the single and multi-point mutations in terms of RP and RMSE

From: An interpretable deep geometric learning model to predict the effects of mutations on protein–protein interactions using large-scale protein language model

Method

Single Mutation

Multi-Point Mutation

S2648

S3421

S4169

M1101

M1707

Rp

RMSE

Rp

RMSE

Rp

RMSE

Rp

RMSE

Rp

RMSE

GES_PPI

0.649

1.132

0.717

1.641

0.689

1.563

0.569

1.762

0.754

2.142

gnn_PPI*

0.627

1.220

0.673

1.694

0.625

1.570

0.543

1.761

0.740

2.176

ProS_GNN

0.608

1.207

0.663

1.724

0.657

1.619

0.510

2.024

0.694

2.343

GeoPPI

0.584

1.188

0.682

1.701

0.592

1.588

0.562

1.885

0.727

2.235

MutaBind2

0.532

1.228

0.690

1.695

0.628

1.596

0.529

1.794

0.712

2.256

TopGBT

0.451

1.352

0.543

1.846

0.409

1.628

–

–

–

–

  1. The bold values represent the best results within each column, corresponding to the specific dataset and evaluation metric
  2. -: the dash sign indicates the results of the corresponding methods are not available
  3. *: gnn_PPI represents the method without the ESM pre-trained model
Back to article page

Journal of Cheminformatics

ISSN: 1758-2946

Contact us