| ||Format||Pages||Price|| |
|PDF (304K)||18||$25||  ADD TO CART|
|Complete Source PDF (9.1M)||546||$70||  ADD TO CART|
The practice of reducing a dose-response test into a single number such as LC50 is convenient but inadequate for distinguishing between dissimilar modes of biological response. When the mean curves of two groups of tests with dissimilar slopes at LC50 cross over at a 50% line, they give rise to equal LC50 cross over at a 50% line, they give rise to equal LC50 numbers for both groups of tests, and if used exclusively to represent the results, the single LC50 will mask this difference. The sensitivity of an organism to an incremental change in the dose is measured by the slope of the dose-response curve at any dose level. The slope for some modes of toxic response can be initially zero at small dose, then increase substantially at higher values, and for other modes it can be infinitely large initially, then level off at higher values of the dose. The management of different chemicals requires different strategies that can be supported with a more comprehensive analysis of existing test data.
Several similar or replicate tests produce slightly different dose-response curves due to uncontrollable experimental and biological test factors. Continued focus on summarizing these tests by averaging their LC50 numbers hinders progress toward discovering a biological response mode from the dose-response points of similar tests.
The paper presents an approach for aggregating a group of similar test data into a single mean dose-response curve with error bounds quantifying the variability of the original data. Groups of similar tests are then classified using generalized exponential dose-response functions whose constants define modes of biological response and facilitate calculation of toxicity and sensitivity. Groups of similar data and classes of different groups of tests are used to demonstrate the application and the utility of the approach.
comparative toxicology, bioassay end points, toxicology scalers, centroid, data aggregation
Senior research scientist, Environmental Protection Agency, Corvallis, OR
Associate scientist, Northrop Services, Corvallis, OR