QSAR model for blood-brain barrier permeation.

Area of Science:

• Computational chemistry
• Drug discovery
• Neuroscience

Background:

• Predicting blood-brain barrier (BBB) permeability is vital for developing neurotherapeutics.
• Experimental evaluation of all potential drug candidates is infeasible.
• Computational models offer a necessary alternative for assessing BBB partitioning.

Purpose of the Study:

• To evaluate the CORAL software for quantitative structure-activity relationship (QSAR) modeling of BBB permeation.
• To establish a computational tool for predicting BBB permeability of novel compounds.

Main Methods:

• Utilized the Monte Carlo technique to develop predictive models.
• Selected correlation weights for molecular features to establish structure-endpoint relationships.
• Employed the CORAL software for QSAR analysis.

Main Results:

• Developed predictive models with good performance across three random data splits.
• The best external validation model achieved a determination coefficient (R²) of 0.896.
• The root mean squared error (RMSE) for the best model was 0.175 with 41 compounds.

Conclusions:

• The CORAL software and the described approach are suitable for predicting blood-brain barrier permeation.
• This method provides a valuable tool for drug discovery in neurotherapeutics.