ToxCML
Why ToxCML?
AI-driven QSAR and Read-Across modeling for fast, ethical, and reliable chemical toxicity screening.
High-Throughput Screening
A high-throughput screening environment that can efficiently evaluate thousands of compounds through a streamlined analysis workflow.
Lower Development Costs
Reduces dependence on costly in vivo and in vitro experiments, lowering overall development costs and accelerating decisions in early-stage screening.
Ethical & Compliant
Supports ethical research practices by minimizing animal testing and complying with internationally recognized research standards.
General Workflow of ToxCML
1. Data Input
Chemical structures entered via SMILES notation, with direct visualization within the software interface.
2. Feature Extraction
For QSAR (machine learning) models: Fingerprints: Morgan Circular, MACCS Keys, Atom Pair Fingerprint (APF). Descriptors: physicochemical properties.
For Read-Across models: Fingerprints: Morgan Circular, MACCS Keys, Atom Pair Fingerprint (APF), RDKit Fingerprint.
3. Machine Learning & Read-Across
Machine learning models: Random Forest, XGBoost, SVM (QSAR-based AI)
Read-Across approach: predicts toxicity based on structurally similar compounds using Tanimoto similarity
Consensus Prediction: combines ML and Read-Across results for reliable, explainable toxicity predictions
4. Assessment & Prediction
14 key toxicity endpoints. Local toxicity: skin sensitization, respiratory irritation, eye irritation, skin irritation. Systemic toxicity: AMES mutagenicity, acute inhalation toxicity, carcinogenicity, drug-induced liver injury (DILI), hepatotoxicity, developmental toxicity, cardiotoxicity, oral toxicity, dermal toxicity, FDA maximum daily dose (FDA MDD).
Result download: save prediction results locally for further analysis and documentation.
Technical Highlights
Try ToxCML free. No sign-up required.
Run a real AI-driven toxicity prediction as a guest, then unlock the full platform whenever you're ready.
Try Tox-CML now