In silico renal clearance model using classical Volsurf approach

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacokinetics

Background:

• Renal clearance is a critical factor influencing drug bioavailability and efficacy.
• Predicting renal clearance aids in drug design and development.
• Diverse compound datasets are essential for robust pharmacokinetic modeling.

Purpose of the Study:

• To develop and validate a quantitative structure-activity relationship (QSAR) model for predicting renal clearance.
• To identify key molecular descriptors that influence renal clearance.
• To classify compounds into high- and low-renal clearance categories.

Main Methods:

• Utilized a dataset of 130 diverse compounds, including central nervous system (CNS) and non-CNS drugs.
• Employed the Volsurf approach for molecular descriptor generation.
• Applied principal component analysis (PCA) and partial least-squares (PLS) regression for modeling.
• Used SIMCA and recursive partitioning for classification.
• Validated models with an external test set of 20 compounds.

Main Results:

• Score plots from PCA and PLS successfully separated high-clearance from low-clearance compounds.
• PLS models demonstrated predictive accuracy for renal clearance.
• Key descriptors influencing renal clearance were identified using PLS coefficient plots, Volsurf profiles, 3D Grid maps, and RP decision trees.
• Topological descriptors (e.g., Molconn-Z) were also evaluated.

Conclusions:

• The developed Volsurf-based models are effective for classifying and predicting renal clearance of unknown compounds.
• Accurate prediction of renal clearance is crucial for understanding drug bioavailability.
• These models serve as valuable tools in early-stage drug discovery and development.