Home Healthcare Scientists Leverage AI in Predicting Side Effects from Combining Drugs

Scientists Leverage AI in Predicting Side Effects from Combining Drugs

Artificial intelligence may just be the answer to various critical issues.

In fact, continuous research and development in this field have proven quite beneficial, judging from the recent accomplishment reports coming from different industries.

For instance, a team from Stanford University is currently utilizing a new artificial intelligence system to better forecast potential side effects, mainly from drug combination.

The new AI-based system is an incredible achievement, considering the known 125 billion possible side effects between all potential drug pairs that make the predicting of how a patient may react to a new drug a risky guessing task.

Dubbed Decagon, the system can help doctors, especially when prescribing drugs to patients who are already under another combined list of medication. What’s more, it can allow researchers to find better drug combinations to treat complex ailments.

Since there are about 5, 000 drugs currently on the market and 1,000 known side effects, determining the effects of different drug combinations is a daunting task.

Marinka Zitnik, a computer science postdoctoral fellow, emphasized this point by saying that it is practically difficult to test a newly-developed drug in combination with other existing drugs. For a single drug test, that would translate to 5,000 new experiments.

The team of researchers at Stanford University started by studying how drugs impact the body’s underlying cellular machinery. They also composed a massive network that describes how over 19,000 proteins interact with each other as well as how different drugs impact such proteins.

MORE – Top 10 Ways Artificial Intelligence is Impacting Healthcare

MORE – Data Science – 8 Powerful Applications

In fact, the team utilized over four million out of the 125 billion known relations between side effects and drugs to come up with a deep learning system that spots patterns in how side effects emerge based on how drugs target various proteins.

The deep learning system created by the Stanford University team looks at patterns regarding drug interaction side effects before forecasting what the consequences of taking multiple drugs in one go would be for a user.

On completion, the researchers would then embark on another step to determine whether the predictions made were accurate. For distance, the artificial intelligence system predicted that atorvastatin, a cholesterol drug, would cause muscle inflammation when combined with amlodipine, a blood pressure drug, despite no previous indication of such a side effect.

The team confirmed this finding with a case report from 2017, which indicated that the combination of the two drugs led to a severe type of muscle inflammation.

Aside from the dangerous drug combination causing muscle inflammation, the researchers discovered at least 10 side effects forecasted by Decagon that were not present in the original data, whereby half of them were recently confirmed.

Jure Leskovec, a computer science associate professor, said that it was amazing how protein interaction networks disclose a lot about drug side effects.

The next plan for the researchers is to improve their work and include additional complex regimens while building more user-friendly tools for physicians.

Furthermore, the CDC revealed that 23 percent of all Americans took a minimum of two prescription drugs in the past month while 39 percent of individuals above 65 in the United States took five or even more.

Source RDMag

Subscribe to our newsletter

Signup today for free and be the first to get notified on the latest news and insights on artificial intelligence

KC Cheung
KC Cheung
KC Cheung has over 18 years experience in the technology industry including media, payments, and software and has a keen interest in artificial intelligence, machine learning, deep learning, neural networks and its applications in business. Over the years he has worked with some of the leading technology companies, building and growing dynamic teams in a fast moving international environment.
- Advertisment -


AI Model Development isn’t the End; it’s the Beginning

AI model development isn’t the end; it’s the beginning. Like children, successful models need continuous nurturing and monitoring throughout their lifecycle. Parenting is exhilarating and, if...