Predicting Blood⁻Brain Barrier Permeability of Marine-Derived Kinase Inhibitors Using Ensemble Classifiers Reveals...

Area of Science:

• Medicinal Chemistry
• Neuroscience
• Drug Discovery

Background:

• Small-molecule kinase inhibitors are successful anticancer drugs.
• Limited success in treating central nervous system (CNS) disorders due to poor blood-brain barrier (BBB) permeability.
• Marine biodiscovery has yielded numerous kinase inhibitors, but few assessed for BBB penetration.

Purpose of the Study:

• To predict the blood-brain barrier (BBB) permeability of marine-derived kinase inhibitors.
• To identify promising marine natural products for CNS drug development.

Main Methods:

• Utilized open-source chemoinformatics and machine learning algorithms.
• Trained models on 332 known CNS-penetrant small molecules.
• Evaluated regression and classification models, including random forest, gradient boosting, and logistic regression.

Main Results:

• Optimized classifiers achieved high accuracy (80%-82% cross-validated, 78%-80% external testing).
• Identified 13 marine-derived kinase inhibitors with suitable physicochemical properties for BBB permeability.

Conclusions:

• Machine learning models can effectively predict BBB permeability in marine kinase inhibitors.
• This approach aids in identifying novel drug candidates for CNS disorders.