Computer models developed that rival animal toxicity testing

[Michael Oberdorfer]

Recent research published in Toxicological Sciences and cited in Nature News reports that a new algorithm developed by Thomas Hartung’s lab at Johns Hopkins can predict toxicity outcomes for thousands of chemicals rivaling current animal toxicity tests.
The work involved the development of a large database of thousands of chemicals based on 800,000 animal tests.
The researchers state that this can vastly improve toxicological testing in the future by replacing expensive animal testing, but “animal tests are harder to replace when it comes to assessing more complex harms, such as whether a chemical will cause cancer or interfere with fertility."