Related Experiment Video
Updated: Jun 10, 2026
13:22
Kinase Inhibitor Screening In Self-assembled Human Protein Microarrays
Published on: October 23, 2019
Virtual screening of selective multitarget kinase inhibitors by combinatorial support vector machines
1Bioinformatics and Drug Design Group, Department of Pharmacy, Centre for Computational Science and Engineering, National University of Singapore, Blk S16, Level 8, 3 Science Drive 2, Singapore 117543.
Molecular Pharmaceutics
|August 18, 2010
Summary
Combinatorial support vector machines (C-SVM) effectively identify dual-target kinase inhibitors for cancer therapy. This virtual screening tool shows promise for discovering multitarget agents with improved efficacy and reduced toxicity.
Area of Science:
- Computational Chemistry
- Drug Discovery
- Bioinformatics
Background:
- Multitarget agents are increasingly explored to enhance therapeutic efficacy and minimize off-target toxicities in drug development.
- Efficient virtual screening (VS) tools are crucial for identifying selective multitarget agents against multiple biological targets.
Purpose of the Study:
- To evaluate Combinatorial Support Vector Machines (C-SVM) as VS tools for discovering dual-inhibitors targeting combinations of nine anticancer kinases.
- To assess the selectivity and false-hit rates of C-SVM in identifying multitarget agents.
Main Methods:
- C-SVM models were trained using known non-dual-inhibitors for 11 combinations of nine anticancer kinases (EGFR, VEGFR, PDGFR, Src, FGFR, Lck, CDK1, CDK2, GSK3).
- The performance of C-SVM was evaluated by its ability to identify known intra- and inter-kinase-group dual-inhibitors.
- C-SVM was compared against other VS methods (Surflex-Dock, DOCK Blaster, kNN, PNN) using kinase inhibitor datasets and the Zinc clean-leads database.
Main Results:
- C-SVM correctly identified 26.8%-57.3% of intra-kinase-group dual-inhibitors and 12.2% of inter-kinase-group dual-inhibitors.
- The method demonstrated fair selectivity, with false-hit rates below 20% for non-dual-inhibitors of the same kinase pairs and low false-hit rates across large compound libraries.
- C-SVM showed comparable dual-inhibitor yields and significantly lower false-hit rates on the Zinc clean-leads dataset compared to other VS methods.
Conclusions:
- C-SVM shows significant potential as a VS tool for the efficient and selective discovery of multitarget agents, particularly for intra-kinase-group targets.
- This approach can aid in developing novel anticancer therapeutics with improved efficacy and reduced side effects.
- The study highlights the utility of C-SVM in navigating complex multitarget drug discovery landscapes.
