Researchers have developed a computer model that could help predict adverse drug reactions, according to a recent study published in the journal Nature, FierceHealthIT reports (Hall, FierceHealthIT, 6/12).
Researchers from the University of California-San Francisco and the Novartis Institutes for BioMedical Research in Cambridge, Mass., created a database of 73 proteins to test whether a library of 656 approved drugs were likely to interact and cause an adverse side effect.
The researchers made more than 1,000 predictions, which they verified through both research and laboratory testing. Approximately 50% of the predictions were correct, and 151 interactions were previously unknown, according to the study (Johnson, Boston Globe, 6/11).
The computer model could help drugmakers identify potential adverse interactions earlier in the process, which could reduce costs.
In addition, the model could help drugmakers identify ways to repurpose drugs to treat other conditions (FierceHealthIT, 6/12).