Prediction
Toxophores
Confident predictions that are fully explainable.
Predict toxic liabilities using Machine Learning (ML) models that are fully explainable.
A module within the MCPairs platform.
How Toxophores Works
Prediction in four steps
Input Compound
Submit a compound for prediction.
Choose Models
Select/deselect models you wish to screen against.
View Results
View your results in Quick or Detailed view.
Export Results
Export results and continue your workflow in MCPairs.
Explore Toxophores
Prediction at your fingertips
Binding Heatmap
Use our binding heatmap to design your way out of off-target activity. Highlight the good and bad parts of the molecule to focus design.
Explainable AI
See the original features and data to understand the predicition.
Clarity
In and out of domain is clearly shown.
Export Results
PowerPoint export to share with colleagues.
Why use Toxophores?
Key Benefits
Fully Explainable
Uses Machine Learning (ML) models that are fully explainable.
In-depth Understanding
Highlight potential pitfalls in lead identification and/or lead optimisation.
Confidence
Output from two models provides higher confidence and links to the original compounds and data.
Web-based access
Access Toxophores through MCPairs Online —no software installation required.
