CardioSim
Tox-CML
An AI-powered in silico chemical screening platform that predicts toxicity across 18 critical endpoints, fast, accurate, and animal-free. No account needed to try it.
Predict Toxicity Without the Wait
Tox-CML combines machine learning with read-across chemistry to deliver fast, explainable toxicity predictions, which reduces cost, time, and reliance on animal testing.
Rapid Screening
Efficiently evaluates thousands of chemical compounds in a single run.
Cost Reduction
Minimizes reliance on expensive in vivo or in vitro testing.
Ethical Compliance
Reduces the need for animal testing throughout early-stage screening.
Resource Efficiency
Prioritizes high-potential candidates for experimental follow-up.
Comprehensive Risk Assessment
Predicts a wide array of toxicological endpoints with explainable AI.
From SMILES to Multi-Endpoint Prediction
The general workflow of Tox-CML, from chemical structure input through to explainable, multi-endpoint toxicity predictions.
Data Input
Accepts SMILES notation for chemical structures and visualizes them directly within the platform.
Feature Extraction
For QSAR models: Morgan Circular, MACCS Keys and Atom Pair fingerprints plus physicochemical descriptors. For Read-Across: the same fingerprints plus RDKit fingerprints.
Machine Learning & Read-Across
Random Forest, XGBoost and SVM (QSAR-based AI); Tanimoto similarity read-across; and a hybrid mfCoQ-RASAR model combining both for robust, explainable predictions.
Evaluation & Prediction
Applicability Domain Analysis, a Fragment Contribution Map for interpretable results, and exportable prediction outputs for reporting.
18 Toxicity Endpoints, One Platform
A single Tox-CML run screens a compound across the following endpoints.
Try Tox-CML free. No sign-up required.
Run a real toxicity screening with Tox-CML right now. Sign in whenever you're ready to unlock the other 6 CARDIOSIM modules.