Enzyme classification integrating LSTM and Prot-BERT sequence encoding

doi:10.1016/j.asoc.2025.113774

Enzyme classification integrating LSTM and Prot-BERT sequence encoding

2025-08-19

0

PRE

AI

DOI：10.1016/j.asoc.2025.113774

Original

Original request for help

Save

Abstract

En 中文

• ECiLPSE, a novel enzyme classification tool combining Prot-BERT and BiLSTM, efficiently predicts 1991 enzyme classes with high accuracy. • Prot-BERT encoding outperforms numeric encoding by capturing intricate details for enhanced predictive accuracy. • The bidirectional LSTM model achieved accuracy (99.14 % training, 98.41 % test) by leveraging Prot-BERT for enzyme classification. • ECiLPSE is available as a web server with user-friendly interface and standalone tool, offering multiple input options for researchers.

Keywords:

ECiLPSE

Prot-BERT

BiLSTM

enzyme classification

web server

Enzyme classification integrating LSTM and Prot-BERT sequence encoding

Abstract

Journal

Organization