Prediction of mutagenicity and analysis of structure basis of 34 drug candidates
Zhong Zhi Fu · Zhongguo yaolixue yu dulixue zazhi · 1992
The CASE (computer-automated structure evaluation) methodology in VAX II was developed by Rosenkranz et al. Recently, we established an expert system of CASE in microcomputers to identify structure moieties responsible for mutagenicity (Ames test). CASE selects its own descriptors from a learning set composed of active and inactive molecules. The descriptors consist of either activated or inactivated fragments. When the structure of an unknown drug candidate enters the computer, all the possible fragments ranging from 2 to 10 atoms accompanied by their hydrogens will be generated and these compared with the previously identified descriptors. On the basis of the presence and / or absence of these descriptors, CASE predicts mutagenicity of the drug candidates of unknown activity.Genetic toxicology tests have become an important factor in the safety assessment of the new drugs and chemicals. All the drug candidates which induce positive responses in genetic toxicology tests can not be applied, although these have shown much more effective. The expert system of CASE may provide the re-sults of the tests of the drug candidates before they were synthesized , so that CASE is a predictive screening method.The purpose of this paper is to report the results of CASE prediction of 34 drugs or potential drugs in the Salmonella / microsome (Ames) test. The activities of mutagenicity of these drug candidates in short-term screening tests have been reported by Aaron, C.S.. The CASE identified 16 fragments which accounted for the mutagenicity, or showed a lack, thereof, of most of the drug candidates. The mutagenicity of the two of 34 drug candidates has not been reported by Aaron; CASE predicted the positive response. The prediction of the other 32 drugs or potential drugs showed the same result as reported by Aaron. The specificity was unexpectedly high (96.4%). This result suggests the effectivity of the CASE expert system for the prediction of mutagenicity.