Rule based approach for prediction of rabbit eye and skin irritation

TOXI 106

Paulius J. Jurgutis, jurgutis@ap-algorithms.com1, Remigijus Didziapetris1, and Pranas Japertas, jurgutis@ap-algorithms.com2. (1) Pharma Algorithms, Inc, 591 Indian Road, Toronto, ON, Canada, (2) Pharma Algorithms Inc, 591 Indian Rd., Toronto, ON M6P 2C4, Canada
The potential of a compound to cause eye or skin irritation is a measure of harmfulness of any drug or chemical intended for topical application. The Draize test on rabbits has long been an industry standard and substantial experimental data is exists. The Draize test involves subjective judgments resulting in inconsistencies and contradictions. Here we present a rule based attempt to model these two very important phenomena. Eye and skin irritation data from nearly 100 references, including ECB-ESIS and RTECS, was critically evaluated and conditions, individual effects and results were documented. The compiled knowledge base consisting of more than 2100 compounds was used in the modeling of the eye and skin irritation. Physicochemical and constructed fragmental descriptors were modeled using binomial PLS method. Results provide rabbit eye and skin irritation prediction models with overall accuracy of 77.6 and 73.6 percent respectively.
 

Poster Session and Awards
6:00 PM-10:00 PM, Tuesday, August 21, 2007 BCEC -- 204 A/B, Poster

Division of Chemical Toxicology

The 234th ACS National Meeting, Boston, MA, August 19-23, 2007