||Health Effects Research Lab., Research Triangle Park, NC. ;Leiden Rijksuniversiteit (Netherlands). ;Lawrence Livermore National Lab., CA. ;Hazleton Labs. America, Inc., Vienna, VA.;Department of Energy, Washington, DC.
Short-term testing methods, encompassing a wide variety of species and genetic mechanisms, have been developed to detect genetic effects produced by chemicals. The use of multiplicity of tests has created a difficult and controversial challenge in the interpretation of mixed test results. The driving principle of this report and the International Commission for Protection against Environmental Mutagens and Carcinogens (ICPEMC) Committee 1 was to combine the major parameters of testing (dose, metabolic activation, sign of response, and the replication of test data) into a single score which could be pooled by entry, by test (e.g., Salmonella reverse mutation), by test class (e.g., bacterial mutation), and by family (in vitro or in vivo), into a composite score for a chemical. The system is designed: (i) to cope with redundant data, disagreement, and sporadically filled matrices; (ii) to supply statistical properties by chemical and by test; and (iii) to have features of self-learning to improve predictive performance and internal consistency for any one of several types of genetic hazard.