QSAR-based permeability model for drug-like compounds.

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacokinetics

Background:

• Intestinal drug permeability is crucial for drug discovery and development.
• Accurate prediction of this parameter aids in selecting viable drug candidates.
• Existing methods may lack efficiency or broad applicability.

Purpose of the Study:

• To develop a Quantitative Structure-Activity Relationship (QSAR) model for predicting intestinal drug permeability.
• To establish a reliable and computationally efficient tool for drug discovery.
• To validate the model's performance using established datasets and experimental data.

Main Methods:

• A database of compounds with Caco-2 permeability values was curated.
• Drug-like properties were used to define chemical space.
• QSAR regression models were developed and optimized.
• Model validation included training/validation subsets and external drug datasets.

Main Results:

• The developed QSAR model demonstrated high accuracy in predicting Caco-2 permeability.
• The model performed well on internal validation sets, including FDA Biopharmaceutics Classification System (BCS) compounds.
• External validation using experimentally determined values for 21 drugs confirmed the model's reliability.
• The model utilizes simple, easily calculable descriptors.

Conclusions:

• The QSAR model provides a reliable method for predicting intestinal drug permeability.
• Its simplicity and accuracy make it suitable for large-scale in silico screening.
• This tool can significantly accelerate the early stages of drug discovery.