Chemical intelligence.
Regulatory certainty.
AI to prioritize candidate molecules, anticipate toxicological risk, and deliver audit-ready evidence.
Supported by premier scientific research funding & innovation programs
Decision Gates
De-risking candidate molecules
at every critical stage gate.
EndoXplain deploys upstream of wet-lab synthesis, transforming toxicological uncertainty into auditable regulatory evidence.
Candidate Triage
Screen virtual chemical libraries and analogue series in silico. Anticipate nuclear receptor affinities, mutagenic structural alerts, and off-target liabilities before committing bench synthesis.
to profile molecular liabilities across candidate cohorts.
Wet-Lab Optimization
Eliminate toxic liabilities prior to experimental commitments. Direct wet-lab capital strictly toward high-confidence leads across OECD TG 455 panels and cruelty-free NAMs.
concentrating wet-lab spend exclusively on de-risked leads.
Submission Intelligence
Bridge computational pharmacology with evidentiary requirements. Generate cryptographic, auditable dossiers with structural SHAP attributions formatted for ANVISA, FDA, and OECD review.
traceable evidence dossiers structured for agency review.
Thesis
Auditable by design.
Built for regulatory defense.
Every prediction comes with mechanistic rationale, so regulators can inspect the structural evidence, not just an opaque score.
Sectors
Toxicology for highly
regulated chemistry.
From pharmaceutical lead candidates to crop protection, the same explainable engine adapted to each regulatory pathway.
ANVISA · FDA
Pharmaceuticals & APIs
Early nuclear receptor profiling to flag endocrine liabilities before wet-lab commitment.
Explore pathway
MAPA · IBAMA · ANVISA
Agrochemicals & Crop Protection
Evidence linking active ingredient safety, environmental persistence, and residue risk.
Explore pathway
OECD NAMs
Cosmetics & Personal Care
Animal-free prediction of skin sensitization and endocrine activity from chemical structure.
Explore pathway
ANVISA · EFSA TTC
Food Contact & Contaminants
Toxicological assessment of non-intentionally added substances (NIAS) and process contaminants.
Explore pathway
OECD Guidelines
Novel Scaffolds & Ingredients
Safety triage for new chemical entities and functional ingredients prior to synthesis.
Explore pathway
Platform / EndoXplain
From molecular hypothesis to audit-ready evidence.
One computational decision engine connecting molecular chemistry and regulatory toxicology, replacing isolated QSAR tools with explainable, auditable models.
Decision co-development
Predictive workflows calibrated to your team’s internal chemistry milestones and screening criteria, replacing generic black-box templates with auditable decision logic.
Multimodal molecular modeling
2D graph representations, 3D receptor binding pocket geometries, and historical bioassay distributions unified into a single coherent screening engine.
Substructure liability flagging
Every liability is resolved to specific chemical pharmacophores and pocket interactions using SHAP feature heatmaps, revealing the mechanistic rationale before wet-lab commitment.
Traceable audit dossiers
Calibrated uncertainty bounds and exportable audit dossiers formatted to meet OECD, ANVISA, FDA, and EPA requirements for direct regulatory submission.
Security & IP
Confidential chemistry
stays confidential.
Architected for preclinical pipelines where a single novel scaffold represents years of capital and patent exclusivity—guaranteeing security directly through the integration methods we provide.
Strict IP isolation & zero retention
Client structures are processed in isolated memory states. Proprietary molecules transmitted via our integrations are never stored or used to train shared models.
Security through integration methods
End-to-end encrypted API connectors (mTLS), scoped cryptographic tokens, and isolated VPC endpoints that integrate directly into your R&D workflow without exposing raw chemical structures.
Cryptographic audit logging
Version-locked model weights, training distributions, and immutable execution logs generated across every integrated toxicological prediction.
Research & Updates
Institutional science.
Regulatory momentum.
Announcements, computational assay validations, and regulatory methodology briefs from the Iterchem research team.
Iterchem secures public research & innovation funding to advance EndoXplain
Backed by FINEP, FAPEG, CNPq, and MCTI, Iterchem is advancing EndoXplain to bridge predictive computational toxicology and operational R&D decision-making.
Beyond black-box QSAR: Explainable AI and uncertainty in endocrine screening
Why standalone probability scores fail in regulatory toxicology—and how combining structural explainability with explicit uncertainty improves candidate triage.
New Approach Methodologies (NAMs): Connecting predictive models to regulatory workflows
How auditable in silico toxicology supports animal-free safety substantiation across cosmetics, agrochemicals, packaging, and pharmaceutical discovery.