Early prediction of drug metabolism and toxicity: systems biology approach and modeling

Area of Science:

• Drug discovery and development
• Computational toxicology
• Pharmacokinetics

Background:

• Many drug candidates fail in clinical trials due to unforeseen human metabolism issues like toxicity and poor pharmacokinetics.
• Early pre-clinical identification of problematic compounds is essential but currently challenging.
• Existing resources include structure-activity relationship (SAR) models, experimental response databases, and systems biology databases.

Purpose of the Study:

• To review the advantages and disadvantages of current computational and experimental resources for predicting drug metabolism.
• To propose directions for integrating these resources to improve early drug candidate elimination.

Main Methods:

• Review of existing literature on computational models (SAR), high-throughput experimental data (pattern databases), and systems biology databases.
• Analysis of the strengths and weaknesses of each resource category for predicting drug metabolism and toxicity.
• Identification of potential integration strategies for these diverse data types.

Main Results:

• Structure-activity relationship (SAR) models offer structure-based predictions but may lack experimental validation.
• 'Pattern' databases provide experimental insights into tissue/organ responses but can be limited in scope.
• 'Systems biology' databases offer pathway and network information but require integration with compound-specific data.

Conclusions:

• No single current resource is sufficient for confident early elimination of problematic drug candidates.
• Integrating SAR models, experimental data, and systems biology approaches is key to improving pre-clinical drug assessment.
• Future directions should focus on developing hybrid systems that leverage the complementary strengths of diverse data sources.