PredAOT: a computational framework for prediction of acute oral toxicity based on multiple random forest models

Area of Science:

• Computational chemistry
• Drug discovery and development
• Toxicology

Background:

• Acute oral toxicity is a key factor in drug candidate failure.
• Early-stage toxicity evaluation is limited by cost and time.
• Predicting toxicity is essential for efficient drug development.

Purpose of the Study:

• To develop a computational framework for predicting acute oral toxicity.
• To enable accurate toxicity prediction for small compounds in mice and rats.
• To support early-stage drug discovery by providing rapid toxicity assessments.

Main Methods:

• Development of PredAOT, a computational framework utilizing multiple random forest models.
• Training models on extensive datasets of 6226 compounds for mice and 6238 for rats.
• Simultaneous prediction of acute oral toxicity in both species.

Main Results:

• PredAOT demonstrates high accuracy in predicting acute oral toxicity.
• The framework offers simultaneous prediction for mice and rats.
• Prediction performance is comparable to or exceeds existing toxicity prediction tools.

Conclusions:

• PredAOT serves as a valuable tool for predicting small compound oral toxicity.
• The framework facilitates quick and accurate toxicity assessments during drug development.
• PredAOT supports efficient and reliable drug discovery processes.