Predicting Drug Toxicity from Integrated Data and Services on the Life Science Semantic Web

Modern drug discovery requires access to machine-understandable data that can be searched, retrieved, and subsequently analyzed using a wide array of analytical software and services.

In this presentation, I will demonstrate the prediction of chemical toxicity through a set of SADI-compliant semantic web services that i) retrieve data from Bio2RDF, the largest linked open data project for the life sciences, ii) compute chemical attributes using the open source Chemical Development Kit (CDK), and iii) ultimately make predictions using formalized (machine understandable) decision trees.

I will show how one can make use of the SADI-enabled SHARE client to automatically identify and invoke the necessary chemical semantic web services by reasoning about formalized service inputs and outputs described using basic types and relations from OWL the Semanticscience Integrated Ontology (SIO) and descriptors from the Chemical Information Ontology (CHEMINF)