Tox Suite is a collection of prediction modules that offer insightful, structure-based calculations of toxicity. It helps to reduce attrition rates of molecular entities that are unlikely to succeed as drug candidates. The toxicity profile can also help direct new compound synthesis and focus animal testing requirements. It allows users to predict various toxicity endpoints from structure, including acute toxicity, aquatic toxicity, endocrine system disruption, mutagenicity, adverse effects on organs, hERG inhibition, and irritation. It also predicts physicochemical properties such as logP, logD, pKa, and aqueous solubility. The software allows users to assess the reliability of predicted values, search an internal library of experimental data, and train prediction models with experimental data to better reflect novel chemical space. Users can also include custom models and in-house prediction algorithms.

Features & Benefits

• Easy to Use
  • Users can simply draw/import a structure for predictions, making it accessible for medicinal, synthetic, and research chemists. No software engineering or programming expertise is required to train the models.
• Fast, Accurate, Reliable Results
  • Quickly calculates properties for single compounds or large libraries, based on curated databases of experimental data. A reliability index, display of similar structures, and literature references are provided to evaluate results.
• Convenient Visualization
  • Visualizes substructure/atomic contributions to a property value with color-mapping on the structure (select modules). Identifies favorable and unfavorable compounds in a library with user-defined color-coding of results in spreadsheet view.
• Deeper Insights
  • Identifies trends and prioritizes compounds easily by creating scatter plots, browsing, filtering, sorting, and ranking libraries of compounds. Provides a complete property profile of molecules in one place for confident decision-making.
• Customizable with In-House Data
  • Users can use their own experimental data to expand the applicability domain of trainable modules, leveraging the accuracy of in-house models within a commercial product.
• Expandable to Third-Party Models
  • Creates a single environment for predicted data by including third-party and in-house models.