148c Self-Organizing Classification of Ecotoxicological and Molecular Information as a Proxy for Identifying Persistent Bioaccumulative and Toxic Substances

Gabriela Espinosa, Departament d'Enginyeria Química, Universitat Rovira i Virgili, Av. dels Països Catalans 26, Tarragona, 43007, Spain, Rallo Robert, Departmament d'Enginyeria Informàtica i Matemàtiques, Universitat Rovira i Virgili, Av. dels Països Catalans 26, Tarragona, 43007, Spain, and Giralt Francesc, Grup de Fenomens de Transport Departament d'Enginyeria Quimica, Universitat Rovira i Virgili, Av. dels Països Catalans 26, Tarragona, 43007, Spain.

The exploration and characterization of the chemical space of REACH compounds is the first step towards the development of new computerized tools which integrate chemical domain knowledge with physicochemical, biological and regulatory information. Environmental risk assessment takes advantage of the so-called proxy indicators which provide reliable information for specific end-points by using simpler and less costly measurements and calculations. In the context of REACH, the development of these proxies requires the establishment of relationships between PBT profiling data and the corresponding molecular information. The identification of the most representative subsets of chemicals within each PBT and vPvB profile contributes to the design and implementation of optimal screening procedures which facilitate the risk assessment process.

In the current work the screening of structural and property profiles for substances of concern such as PBT and vPvB has been performed. The modes of toxic action (MOA) were used as categorical indicators of effects of chemicals in the environment and similar effect patterns were identified by cluster analysis using self organizing maps. Molecular information was added to facilitate the above clustering process on the basis that similar compounds share a common mode of action (chemical reactivity).



Web Page: www.etseq.urv.es/fet/1.html