I-SOLV: a new surface-based empirical model for computing solvation free energies

Area of Science:

• Computational Chemistry
• Physical Chemistry
• Molecular Modeling

Background:

• Accurate calculation of solvation free energies is crucial for understanding chemical processes.
• Existing theoretical models can be computationally intensive.
• Empirical models offer a balance of speed and accuracy.

Purpose of the Study:

• To develop and validate a new empirical model, I-SOLV, for computing solvation free energies.
• To assess the performance of I-SOLV against established solvation models.
• To demonstrate the viability of empirical models as accurate alternatives.

Main Methods:

• Developed the I-SOLV empirical model based on atomic contributions, solvent-accessible surface area, and surface tension.
• Implemented 49 atom types for common elements in organic molecules.
• Parameterized surface tensions using experimental solvation free energy data for 532 molecules.

Main Results:

• I-SOLV achieved a mean unsigned error of 0.39 kcal/mol on a test set of 82 molecules.
• Outperformed widely used Poisson-Boltzmann/Surface Area (PB/SA) and Generalized Born/Surface Area (GB/SA) models.
• Demonstrated high accuracy comparable to more sophisticated theoretical approaches.

Conclusions:

• The I-SOLV model provides a simple and accurate method for calculating solvation free energies.
• Empirical solvation models can be highly effective and competitive with theoretical methods.
• I-SOLV represents a valuable tool for computational chemistry research.