Physics-Based Method for Modeling Passive Membrane Permeability and Translocation Pathways of Bioactive Molecules.

Area of Science:

• Computational chemistry
• Pharmacokinetics
• Drug discovery

Background:

• Accurate assessment of molecular permeability is crucial for drug development.
• Favorable ADME properties are key for selecting viable drug candidates.

Purpose of the Study:

• To develop a novel, physics-based computational method for rapid modeling of passive molecular permeation across lipid membranes.
• To predict drug candidate permeability and inform ADME property assessment.

Main Methods:

• Utilized heterogeneous solubility-diffusion theory with all-atom 3D solute structures.
• Modeled the lipid bilayer using an anisotropic solvent model with dielectric and hydrogen bonding profiles.
• Determined optimal translocation pathways by simulating molecular movement and optimizing orientations within the membrane.
• Calculated membrane-bound states, free energy profiles, and permeability coefficients.

Main Results:

• The method demonstrated high accuracy in predicting permeability coefficients across various experimental models.
• Strong correlations (R² values ranging from 0.52 to 0.88) were observed against pure lipid membranes, PAMPA-DS, BBB, and Caco-2/MDCK assays.
• Root-mean-square errors (RMSE) were within acceptable ranges (0.87 to 1.59 log units).

Conclusions:

• The developed physics-based method provides a fast and accurate computational tool for assessing molecular permeability.
• This method can significantly aid in the early selection of drug candidates with desirable ADME profiles.
• The approach offers a valuable alternative for predicting passive permeation, complementing experimental assays.