In silico ADMET prediction: recent advances, current challenges and future trends

Area of Science:

• Computational chemistry and toxicology
• Pharmacokinetics and drug safety assessment

Background:

• Numerous small molecules (drugs, pollutants) impact health, necessitating robust safety assessments.
• Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) properties are crucial for evaluating compound risks.
• High drug withdrawal rates and the cost/labor of traditional methods drive the need for efficient in silico approaches.

Purpose of the Study:

• To review recent advancements in in silico ADMET prediction.
• To highlight a newly developed substructure pattern recognition method for ADMET prediction.
• To discuss current challenges and future research directions in the field.

Main Methods:

• Review of in silico techniques for ADMET property estimation.
• Emphasis on a novel substructure pattern recognition methodology.
• Discussion of model validation, application domains, and global vs. local models.

Main Results:

• Recent progress in in silico ADMET prediction methods.
• Demonstration of a substructure pattern recognition approach's potential.
• Identification of key challenges including model applicability and validation.

Conclusions:

• In silico ADMET prediction is vital for drug discovery and risk assessment.
• Substructure pattern recognition offers a promising avenue for improved prediction.
• Future research should focus on computational systems toxicology and data integration for systemic prediction.