Significance analysis and multiple pharmacophore models for differentiating P-glycoprotein substrates

Area of Science:

• Pharmacology
• Computational Chemistry
• Drug Discovery

Background:

• P-glycoprotein (Pgp) mediates drug efflux, impacting drug absorption, distribution, and clearance.
• Early identification of Pgp interaction potential is vital for optimizing drug candidates.

Purpose of the Study:

• To develop a robust predictive pharmacophore model to differentiate P-glycoprotein substrates from nonsubstrates.
• To aid in the selection and optimization of drug candidates by assessing Pgp interaction potential.

Main Methods:

• Supervised analysis of three-dimensional (3D) pharmacophores from 163 published compounds.
• Generation of pharmacophores from conformers of Pgp substrates and nonsubstrates.
• Application of a novel pharmacophore-specific t-statistic algorithm and construction of a classification tree using nine distinct pharmacophores.

Main Results:

• A classification tree model accurately distinguished Pgp nonsubstrates from substrates with 87.7% accuracy on the training set and 87.6% on an external test set.
• Each of the nine identified pharmacophores can independently serve as an accurate marker for potential Pgp substrates.

Conclusions:

• The developed pharmacophore model provides a reliable method for predicting P-glycoprotein substrate potential.
• This approach can significantly improve early-stage drug candidate selection and optimization in pharmaceutical research.