TCMFP: a novel herbal formula prediction method based on network target's score integrated with semi-supervised
Qikai Niu1, Hongtao Li2, Lin Tong2
1Institute of Chinese Materia Medica, China Academy of Chinese Medical Sciences, Beijing 100700, China.
Briefings in Bioinformatics
|March 20, 2023
Summary
This study introduces a novel AI-driven approach for predicting effective Traditional Chinese Medicine (TCM) herbal formulas. The method integrates TCM experience with network science to optimize herbal remedies for diseases like Alzheimer's.
Area of Science:
- Computational biology
- Pharmacology
- Traditional Chinese Medicine (TCM)
Background:
- TCM herbal therapy relies heavily on empirical knowledge, hindering systematic discovery of complex multi-target formulas.
- Integrating traditional wisdom with modern pharmacology for disease treatment is challenging.
Purpose of the Study:
- To develop an efficient computational approach (TCMFP) for predicting optimal herbal formulas.
- To combine TCM experience, artificial intelligence, and network science for drug discovery.
Main Methods:
- Developed a herbal formula prediction approach (TCMFP).
- Integrated herb score (Hscore), pair score (Pscore), and formula predictive score (FmapScore).
- Utilized intelligent optimization and genetic algorithms for formula screening.
Main Results:
- Validated Hscore, Pscore, and FmapScore using network analysis.
- Successfully generated herbal formulas for Alzheimer's disease, asthma, and atherosclerosis.
- Functional enrichment analysis confirmed the efficacy of predicted targets.
Conclusions:
- TCMFP offers a novel strategy for optimizing herbal formulas and advancing TCM-based drug development.
- The approach bridges traditional knowledge with modern computational methods for efficient therapeutic discovery.
