Analysis of structure-Caco-2 permeability relationships using a property landscape approach

Area of Science:

• * Medicinal Chemistry and Drug Discovery
• * Computational Chemistry and Cheminformatics

Background:

• * Understanding the relationship between chemical structure and Caco-2 cell permeability is crucial for modern drug discovery.
• * Caco-2 permeability is a key indicator of intestinal drug absorption and bioavailability.

Purpose of the Study:

• * To systematically characterize the Caco-2 permeability landscape of a benchmark dataset of 100 molecules.
• * To explore the influence of molecular representation on predicting Caco-2 permeability.
• * To introduce and analyze
• permeability cliffs
• highlighting structural similarities with significant permeability differences.

Main Methods:

• * Employed property landscape modeling to analyze pairwise comparisons of Caco-2 permeability and chemical structures.
• * Utilized two distinct molecular representations: continuous properties (for QSPR) and various molecular fingerprints.
• * Identified and analyzed
• permeability cliffs
• within the dataset.

Main Results:

• * Observed a clear dependence of the Caco-2 permeability landscape on the chosen molecular representation.
• * Identified
• permeability cliffs
• characterized as
• shallow cliffs
• (less than two log units difference).
• * Demonstrated the utility of property landscape modeling for ADME (Absorption, Distribution, Metabolism, and Excretion) studies.

Conclusions:

• * The molecular representation significantly impacts the characterization of the Caco-2 permeability landscape.
• * The property landscape modeling approach provides valuable insights into structure-permeability relationships.
• * This methodology can be extended to model other ADME-relevant properties for enhanced drug discovery pipelines.