Structure Activity Relationships (SARs) Using a Structurally Diverse Drug Database: Validating Success of Predictor...

Area of Science:

• Pharmacology
• Computational Chemistry
• Drug Discovery

Background:

• ADME/Tox technology is crucial for predicting drug efficacy in discovery.
• Evaluating Food and Drug Administration (FDA) consumer drug profiles and data is essential.
• In silico prediction models require validation with real-world pharmacological data.

Purpose of the Study:

• To validate the prediction models for plasma protein binding (PPB) and bioavailability (BIO) within the KnowItAll platform.
• To assess the utility of an in silico platform for analyzing consumer drug profiles.
• To create a searchable database of pharmaceutical and pharmacological properties for consumer drugs.

Main Methods:

• Collected 14 pharmaceutical and pharmacological properties for 75 diverse consumer prescription drugs from FDA profiles.
• Utilized Bio-Rad's KnowItAll platform with integrated ADME/Tox in silico predictors.
• Extracted data from Portable Document Format (PDF) files of FDA consumer drug profiles.

Main Results:

• The study focused on validating in silico prediction models for PPB and BIO.
• Data completeness varied for some properties across the analyzed drugs.
• The platform was populated with data for 75 structurally diverse consumer drugs.

Conclusions:

• The validation of PPB and BIO prediction models is critical for reliable in silico drug assessment.
• The developed FDA Consumer Drug Database provides a valuable resource for pharmaceutical research.
• Further refinement of in silico models may be necessary based on data completeness observations.