Computational Toxicology Services LLC is based in the Maryland suburbs of Washington D.C. and was started by Joseph F. Contrera, Ph.D. who was formerly the Director of the FDA Center for Drug Evaluation and Research (CDER) Informatics and Computational Safety Analysis Staff (ICSAS) (www.fda.gov/cder/Offices/OPS/default.htm).

Computational Toxicology Services LLC specializes in computer based predictive toxicity screening of pharmaceuticals, food contact substances and industrial and environmental chemicals using MDL-QSAR (quantitative structure activity) software and published FDA database modules (see publication list). Computational Toxicology Services was established to make toxicity screening using FDA published MDL-QSAR models available to the scientific community. MDL-QSAR predictive modules were created for the major toxicology studies necessary for the marketing of pharmaceuticals and food contact substances in the US and internationally. MDL-QSAR computational toxicology models are currently being used as decision support tools for safety assessment by the FDA (1-5).The Computational Toxicology Services modules contain information derived from public sources and available under the freedom of information act. Computational Toxicology Services offers a cost effective solution for pharmaceutical and chemical industry clients with toxicity screening needs that do not justify the investment in QSAR software, database modules and staff required to carry out computational toxicology screening.

Example Computational Toxicology Applications
Pharmaceutical lead selection in drug development (2)
Safety screening and qualification of pharmaceutical contaminants and degredants (3)
Safety screening of drug metabolites
Safety screening of food additive and food contact substances (1, 5)
Hazard and risk prioritization of chemicals

QSAR Testing Limitations
Some classes of chemicals cannot be tested due to limitations of the software or training dataset. Inorganic chemicals (salts and metals), high molecular weight substances (>1200 Daltons), peptides, polysaccharides, proteins, fibers, organometallic chemicals, gases, and complex mixtures of chemicals cannot be modeled. Domain of applicability or coverage; As with all QSAR software, the toxicological properties of test compounds with chemical structures outside of the domain of applicability of MDL-QSAR training data sets cannot be predicted.

Computational toxicology based estimates of toxicity are intended to be used as supplemental decision support information in combination with traditional toxicology testing as part of a weight of evidence approach to hazard and risk assessment.