Session 1: Data Science

OpenTox USA 2015, Session 1: Data Science

Data Science
Session Chair: Barry Hardy, Douglas Connect

Wednesday, 11 February 2015
from 8:45 to 10:45

Speakers Title
Carolyn Mattingly, NCSU Integrating literature-based curated data to predict mechanisms of toxicity
Ann Richard, US EPA Use of Knowledge-informed Chemotypes to Explore the ToxCast/Tox21 Chemical-Data Landscape
Ruili Huang, NCATS, NIH Modeling chemical toxicity - The Tox21 data crowdsourcing challenge 2014
Richard S. Paules, NIEHS, NIH Moving Forward in Phase III of Tox21
Session 1 Summary

Data provides a foundation on which all of our analysis and applications are built, from hypotheses-generation in basic research through to using data as information supporting risk assessment. We need data to be organised in such a way that it is clear in meaning and subject to evaluation. We need to be able to integrate data from multiple sources, to have harmonisation on metadata and processed data representations, to have interoperability across resources.

Finally, we need to be able to organise data as heterogenous evidence which can be mapped to pathways for mechanistic evidence purposes or to decision supporting or regulatory applications with defined information purposes where we need to make best judgements. This session will address the issues raised in putting in place a stronger data science to our underpinning data foundation, addressing both social and technical issues, and providing directions forward for a community effort supporting data science accelerating alternative testing methods development and risk assessment application.