Table 6 Evaluation of ML regressor models constructed using word embeddings derived from PLMs on an independent dataset

From: Prediction of hemolytic peptides and their hemolytic concentration

Embedding source

ML model

R

R2

MAE

MSE

ESM2-t36

ETR

0.706

0.486

0.808

1.105

ESM2-t33

ETR

0.711

0.495

0.786

1.084

ESM2-t6

ETR

0.677

0.452

0.820

1.177

ProtBERT

RFR

0.410

0.115

1.076

1.407

BioBERT

ETR

0.616

0.366

0.927

1.362

ProtBERT+BiLSTM

RFR

0.646

0.449

0.825

1.265

  1. R Pearson correlation coefficient, R2 coefficient of determination, MAE mean absolute error, MSE mean squared error, Bold values indicate the best-performing model.
Back to article page