A major objective of Computational Toxicology is to provide reliable and useful estimates in silico of (potentially) harmful actions of chemicals in humans. Predictive models are commonly based on in vitro and in vivo data, and aims at supporting risk assessment in various areas, including the environmental protection, food, and pharmaceutical sectors. The field is however hampered by the lack of standards, access to high quality data, validated predictive models, as well as means to connect toxicity data to genomics data.
We present a framework and roadmap for a novel public infrastructure for predictive computational toxicology and chemical safety assessment, consisting of: (1) a repository capable of aggregating high quality toxicity data with gene expression data, (2) a repository where scientists can share and download predictive models for chemical safety, and (3) a user-friendly platform which makes the services and resources accessible for the scientific community. Databases under the framework will adhere to open standards and use standardized open exchange formats in order to interoperate with emerging international initiatives, such as the FP7-funded OpenTox and ToxBank projects.
The infrastructure will strengthen and facilitate already ongoing activities within in silico toxicology, open up new possibilities for incorporating genomics data in chemicals safety modeling (toxicogenomics), as well as deepen the exploitation of signal transduction networks. The initiative will lay the foundation needed to boost decision support in risk assessment in a wide range of fields, including drug discovery, food safety, as well as agricultural and ecological safety assessment.
2012. Vol. 211, no Supplm, S59- p.