Every day we come into contact with common pesticides. Which chemicals, in what dosage, can cause us harm? See how the EPA used crowdsourcing to deliver more accurate predictive analysis.
Predictive toxicology research is critical to chemistry testing. Since current chemical testing is expensive and time consuming, only a small fraction of chemicals have been fully evaluated for potential adverse effects.
EPA created a database of 1,800 chemicals with the goal of determining the lowest dose at which a change was observed in animal toxicity tests.
The EPA Toxcast project consisted of 5 algorithmic challenges that ran in sequence to predict minimum toxicity using the EPA database of 1800 chemicals and other publicly available sources.
The winning submission provided a novel approach towards addressing toxicity and has a very high potential for further advancement within the space.
EPA scientists presented the results and findings at the 2014 EPA Toxcast Data Summit.
“We love it.” and “We shared it with the scientists and they were impressed!” – EPA Representative