Using Molecular Initiating Events To Generate 2D Structure-Activity Relationships for Toxicity Screening

Area of Science:

• Computational toxicology
• Chemical informatics
• Drug discovery

Background:

• Molecular initiating events (MIEs) are driven by chemical interactions.
• Understanding chemical properties is key to predicting these interactions.
• Existing methods for characterizing MIEs are limited.

Purpose of the Study:

• To develop computational structure-activity relationships (SARs) for predicting MIEs.
• To identify chemical characteristics associated with interactions at human toxicology targets.
• To create a screening tool for novel chemicals.

Main Methods:

• Utilized chemical informatics to search the ChEMBL database for molecules binding to human toxicology targets.
• Identified common 2D fragments in interacting compounds.
• Developed bespoke 2D structural alerts based on fragment analysis and literature review.
• Performed internal validation of the developed SARs.

Main Results:

• Successfully generated 2D structural alerts for predicting MIEs.
• Validated the quality of the developed SARs through internal testing.
• Demonstrated a novel approach to characterizing chemical interactions relevant to toxicology.

Conclusions:

• The developed SARs represent a foundational tool for screening novel chemicals.
• This work provides a key first step in addressing the challenge of MIE characterization.
• The approach aids in understanding potential human interactions of chemicals.