SBIR-STTR Award

The mutagenicity of chemicals is believed to be an important feature towards the chemical causing cancer. At present, there is a limited amount of feasible computational models to estimate the mutagenic potential of new substances. We proposed to develop a new method for predicting mutagenicity by a combination approach of multiple genicity databases using the M-CASE software. The databases will be separated by different Salmonella typhimurium strains and by test conditions. A similar condition model has previously been developed for carcinogenicity as a guide in this mutagenicity study In addition, we have developed a single preliminary database to elucidate the potential of this project. This proposal will outline the carcinogenicity results and estimate the improvement possible for mutagenicity. In addition, we have developed a single preliminary database to elucidate the potential of this project. For the Phase I project, we wish to develop four to six of these mutagenicity modules with the goal of producing a total of 24 mutagenicity databases after Phase II funding. PROPOSED COMMERCIAL APPLICATIONS: The need for pre-emptive determination of mutagenic compounds is immense to the regulatory, environmental and pharmaceutical and pharmaceutical industries. In the regulatory bodies such as the Food and Drug Administration (FDA) and the Environmental Protection Agency (EPA), the ability to quickly and accurately assess a new chemical for potential harmful activity to human or the environment is critical to an effective screening process. The incorporation of this multiple level analysis for mutagenicity as a tool to assist regulators will increase productivity and accuracy of their work. In the pharmaceutical industry, the ability to locate pharmacophores early in the stages of research and development will potentially save massive amounts of time and money for the company. This preventative assessment of mutagenicity of new chemical is clearly important for companies in the environment industry.

Computational expert systems provide an inexpensive and fast alternative to short term genotoxicity assays such as the Ames test. Validation studies show the predictive capability of the MCASE system is about 85 percent. That is, 85 percent concordance is expected between experiment and computational genotoxicity predictions for new chemicals. The strong correlation between chemical structure and genotoxicity is particularly useful for 'in silico' prescreening of new drugs in the pharmaceutical industry. The new Salmonella database modules being developed in this work will be made available online to the public through the InfoTox web site (http://www.l-tox.com). Additionally, NIH grantees will be allowed unlimited access to the Salmonella modules through InfoTox at no cost. Collaboration will be sought with large drug companies, with mutual exchange of data. Thus the databases will evolve and improve over time as new data are submitted to form a centralized pool of mutagenicity data, that will provide a resource for avoiding unneeded testing of chemicals structurally similarly to those that are already thoroughly understood. Our collaborators at the FDA/CDER will lead the effort to build this industrial consortium.

Thesaurus Terms:
Salmonella typhimurium, artificial intelligence, chemical information system, chemical structure function, computer simulation, computer system design /evaluation, informatics, mathematical model, mutagen testing chemical property, gene mutation, high throughput technology, mutagen, physical property, toxicology alternatives to animals in research, computer program /software, data collection, statistics /biometry