Development and Test of Highly Accurate End Point Free Energy Methods. 4. Expanding Solvents Capability and logBB...

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Accurate prediction of solvation free energies (SFEs) and partition coefficients (logP) is crucial for drug discovery.
• Existing models often have limitations in diverse organic solvent environments.
• Predicting blood-brain barrier (BBB) permeability (logBB) is essential for CNS-targeted drug development.

Purpose of the Study:

• To expand the Poisson-Boltzmann surface area (PBSA) solvation model for high-throughput calculations across diverse organic solvents.
• To develop a predictive model for blood-brain barrier (BBB) permeability (logBB).
• To establish computationally efficient methods for predicting key ADMET descriptors.

Main Methods:

• Parametrization of the solvent accessible surface area (SASA) nonpolar term for 27 organic solvents using 1246 experimental SFE measurements (Minnesota Solvation database).
• Application of the expanded PBSA model for calculating SFEs and logP values.
• Development of a quantitative model for logBB using multivariable linear regression based on PBSA-derived hydration free energies and organic-solvent SFEs.

Main Results:

• The expanded PBSA model achieved an RMSE of 1.10 kcal/mol for SFEs across the MNSol dataset.
• PBSA yielded an RMSE of 2.02 log units for water-organic solvent logP values.
• The developed logBB model demonstrated stable predictive performance with RMSE ≈ 0.65 log units and MAE ≈ 0.5 log units on test sets.

Conclusions:

• The extended PBSA model enhances applicability to a wide range of solvents for calculating SFEs and logP.
• The new logBB model provides a physically motivated and computationally efficient approach for predicting BBB permeability.
• These advancements offer valuable tools for accelerating the prediction of crucial ADMET descriptors in drug discovery.