Powered by In Silico AI predictions
Analyze the toxicity of chemical substances
Evaluate the safety of chemical substances through AI-based analysis
and predict the likelihood of toxicity.
Chemical hazard screening with AOP-based explainable AI
Integrated toxicity mechanism prediction from Molecular Initiating Events (MIE) to Adverse Outcomes (AO) based on ToxCast Assay and OECD Test Guidelines
Reliable Results Through Data-Driven Modeling
Leverages data models trained on US EPA ToxCast (9,000+ chemicals, 1,000+ bioassays) and
OECD Test Guidelines, providing confidence scores and performance metrics for each prediction.
AOP-Based Explainable Toxicity Prediction
Uses AOP-based MIE → KE → AO linkage to analyze disease relevance, derive toxicity priorities,
and provide explainable rationale for high-risk substance selection while supporting test strategy development.
Goes beyond Positive/Negative outcomes
to provide interpretation at the level of biological pathways and regulatory assessment.
Validating Results by Comparing Experimental and Predicted Values
Review model results alongside experimental data to support evidence-based screening decisions,
not just raw predictions.
By reviewing model-specific performance metrics and per-endpoint results together,
you can assess prediction confidence and prioritize follow-up testing.
From Molecular Structure-Based Screening
to Results Report
Molecular Structure-Based In Silico Screening
Automatically computes molecular fingerprints
and simultaneously predicts multiple endpoints using models trained on ToxCast and OECD Test Guidelines.
Explainable AI Prediction Model with AOP
Integrates AO-level models (OECD TestGuideline-based) that directly correspond to regulatory endpoints with MIE/KE-level models (ToxCast-based) that elucidate cellular and molecular mechanisms — all within the AOP framework —
enabling interpretation of mechanisms at the biological pathway level.
Automatic Applicability Domain Validation
Automatically verifies whether a predicted substance falls within the model's training domain,
and provides advance warnings for substances outside that scope.
Step-by-Step Workflow Exploration
Cumulatively tracks analysis history by search condition changes in a branching structure,
allowing comparison of multiple scenarios.
Combines multiple screening actions — toxicity prediction, similar substance search, etc. — with substance search conditions.
Automatic Regulatory Compliance Report Generation
Automatically generates reports in Excel/PDF format
based on cover sheet and per-endpoint prediction results.
With ToxBAI, predict the risk of large-scale chemical substances
and automate everything from high-risk substance screening to report generation.
Want to learn more about Crissit products?
Providing customized AI solutions based on extensive knowledge from documents and databases of organizations/enterprises
Integrated search of global safety data in one environment with AI-powered automation for rapid analysis of key endpoints, efficiently supporting from data collection to report generation