Powerful Self-learning Software Platform

Machine-learning platform offering unique benefits

World's Largest Toxicity Database

ViridisChem database contains most known (more than 90 million) chemicals with over 50 physical, chemical and toxicological properties per chemical, has comprehensive regulatory information from over 130 regulatory lists worldwide, and offers full GHS classification.

The following features enable ViridisChem to offer sophisticated search capability to identify less toxic alternatives based on researcher’s exact project needs:

  • ViridisChem database identifies each chemical as belonging to specific class (e.g. acid, base, neutral, reagent, solvent, VOC, etc.) or functional groups (aldehyde, ketone, carboxyl, etc.), and correlates its acute/chronic health concerns. There are more than 20 different classes and over 50 functional groups.
  • Chemicals’ relevance within specific sector (pesticides, herbicides, polymers, pigments, binders or resins, alpha/beta hydroxy acids, phthalates, parabens, etc.) is identified so that search for chemicals within specific sector can be specified.
  • Most chemicals’ health, safety, and ecological hazard risks are ranked, and its global regulatory concerns are identified. Chemical’s GHS classification, DOT (Dept. of Transportation) codes, NFPA scores are also logged. This information is critical to develop Safety Data Sheets (SDS).

Real-time Toxicity Evaluation

For new/postulated and proprietary chemicals, based on structural and functional properties, ViridisChem platform runs its in-house prediction models to estimate toxicological properties. This information is then used to calculate endpoint scores that are then aggregated to higher-level Ecological Health and Safety Scores. This means that if a scientist draws a structure of new, unknown, or even postulated molecule, ViridisChem platform can offer real-time toxicity predictions. Scientist can then explore if related derivatives are more or less toxic before selecting the target molecule.

Companies can utilize this feature to get toxicity-prediction of the drug-targets during lead-identification to identify or eliminate high-risk drug-targets, or to explore less toxic alternatives even before they spend R&D efforts . By integrating the tool within their R&D system, they can even automate the toxicity prediction for their entire list of targets.

Powerful Machine-learning Technologies

Utilizing “Deep machine-learning neural network” technologies, ViridisChem platform is constantly learning from past toxicity-predictions to offer increasingly accurate results. To do this, ViridisChem has utilized its extensive experimental data repository and its world’s largest highly curated database. As more data becomes available over time, the prediction accuracy will continue to increase.