New AI-Based System to Detect Side Effects of Combination Drugs


AI-based system called as Decagon is developed by Stanford researchers to predict adverse effects of drug combinations

High prevalence of chronic disorders is increasing adoption of several drugs among population. Prescription of more than one drug simultaneously may have adverse effects on health of an individual, which cannot be identified by conventional detections methods. Novel system developed by team of computer scientists at Stanford University is capable of detecting consequences of combining two drugs using AI-driven computer system. The research was published in the journal Bioinformatics in July 2018.

Drug combinations are a remarkably unstudied area, but as Marinka Zitnik explains, “it’s practically impossible to test a new drug in combination with all other drugs, because just for one drug that would be five thousand new experiments.” Around 40% of Americans over the age of 65 consume five or more different drugs on regular basis and doctors often have to monitor patients to see if any of those drugs combine to create adverse side effects.

The team developed deep learning system called Decagon, which is trained on data encompassing over 19,000 proteins and how different drugs interact with those proteins. The system can effectively predict the consequences of combining any two different drugs. To examine the efficiency of system, the team examined 10 of the systems predicted drug pair interactions that didn’t have clearly known adverse interactions. Furthermore, the team is working on making Decagon more user-friendly tool so that doctors can easily navigate for information when prescribing combinations of drugs. Currently, the system only evaluates drug pairs but the researchers hope to expand that into more complex combinations of drugs in near future.


Comments are closed.