Related Experiment Video
Updated: Jun 26, 2026
08:49
Identification of Mediators of T-cell Receptor Signaling via the Screening of Chemical Inhibitor Libraries
Published on: January 22, 2019
Identification of small-molecule TNF-α inhibitor candidates using machine learning-guided screening and multiscale
Fan Liu1, Licai Xu1, Mengliang Cai1
1Department of Rehabilitation Medicine, Hubei No. 3 People's Hospital of Jianghan University, Hubei, China.
Scientific Reports
|June 24, 2026
Summary
This study developed a computational method to find new small-molecule inhibitors for tumor necrosis factor-alpha (TNF-α), a key factor in low back pain from lumbar disc herniation (LDH). The best candidate shows promise for future LDH therapies.
Area of Science:
- Computational chemistry and drug discovery
- Biomedical engineering
- Pharmacology
Background:
- Lumbar disc herniation (LDH) causes chronic low back pain, with tumor necrosis factor-alpha (TNF-α) central to inflammation.
- Existing biologic TNF-α inhibitors have limitations including systemic administration, high cost, and poor disc penetration.
- There is a need for novel, targeted therapeutic strategies for LDH.
Purpose of the Study:
- To develop and apply a multiscale computational framework for identifying novel small-molecule TNF-α inhibitors.
- To screen a large compound library for potential therapeutic agents against LDH.
- To provide mechanistic insights into the binding behavior of identified candidates.
Main Methods:
- Machine learning models (Random Forest) trained on validated TNF-α inhibitors.
- Virtual screening of 61,534 compounds using the trained model.
- In silico evaluation including Glide XP docking, ADMET prediction, molecular dynamics (MD) simulations, MM/GBSA, and DFT analysis.
Main Results:
- A Random Forest model achieved high performance (ROC-AUC = 0.92) for inhibitor prediction.
- The top-ranked compound (8009-0259) demonstrated favorable binding affinity and stable interactions with key residues (Tyr151, Gln61) in MD simulations.
- Computational analyses indicated favorable binding free energy, moderate electronic stability, and potential for intermolecular interactions.
Conclusions:
- The developed computational framework effectively prioritizes small-molecule TNF-α inhibitors.
- The lead compound (8009-0259) shows significant therapeutic potential for LDH.
- Further experimental validation is warranted to confirm the efficacy of identified candidates.
