A genotoxin is an agent that is capable of inducing damage to or heritable changes in genetic material. Such agents are thought to play a significant role in carcinogenesis, and possibly aging and heart disease as well. Scientific research in genetic toxicology aims to identify and understand genotoxins and their possible mechanisms of action. The primary focus of this project is the development of appropriate statistical methodologies for analysis of genotoxicity data in a variety of animal and microbial systems. Investigations were undertaken to develop statistical analyses and designs for experiments assessing chromosome loss in yeast. Data from these biological assays often exhibit a unimodal """"""""umbrella"""""""" pattern in the dose-response, and the statistical analyses were designed to take this possible nonmonotone curvature into account. Minimal design configurations were identified, below which the test procedures lose power to detect true """"""""umbrella"""""""" departures from a no-effect model. These methods are applicable to any toxicological assay where downturns in dose-response are evidenced. Additional research examined quantitative trends in the association between mutagenic and carcinogenic measures oF potency. A significant positive correlation was seen between the two potencies, although the spread of the data cloud was too great for the overall result to be of predictive use. Classification by chemical class further indicated strong positive correlations for nitrosamines and related compounds. These compounds are known to possess strong carcinogenic potential.
