TCM-Navigator: A deep learning workflow for TCM-like compound generation and evaluation

Sep 26, 2025·
Feiying Chen
Feiying Chen
,
Victor Jun Yu Lim
,
Mingyu Li
,
Hao Fan
· 1 min read
TCM-Navigator workflow and evaluation overview.
Abstract
TCM-Navigator is a data-driven and deep learning-based workflow for in-silico generation, quality control, and physics-based evaluation of TCM-like molecules. It combines TCM-Generator, a transfer learning and LSTM-based chemical language model, with TCM-Identifier, an AttentiveFP-based model for recognizing TCM-like characteristics, and downstream target-aware evaluation.
Type
Publication
Briefings in Bioinformatics, 26(5)
Status
Peer-reviewed Open access
publications

Citation. Feiying Chen, Victor Jun Yu Lim, Mingyu Li, and Hao Fan. TCM-Navigator: a deep learning-based workflow for generation and evaluation of traditional Chinese medicine-like compounds for drug development. Briefings in Bioinformatics, 2025.

This work connects molecular generation with practical drug-development filters: a TCM-like chemical language model generates large focused libraries, an AttentiveFP-based identifier scores TCM-like characteristics, and physics-based evaluation helps prioritize target-ligand pairs.

AI-Driven Drug Discovery Molecular Generation Traditional Chinese Medicine
Feiying Chen
Authors
Feiying Chen
PhD Student in Biology
Feiying Chen is a PhD student in Biology at Tsinghua University. His research interests include perturbation biology, AI virtual cell modeling, single-cell transcriptomics, AI-driven drug discovery, and phenotype-based drug design.