Opioid disorders affect around three million Americans every year and 80,000 Americans die from an overdose. When Opioid receptors are activated, it gives the drug user a feeling of euphoria. The addiction the drugs induce weakens the person physically, sometimes leading to death. Traditional treatment methods employing drugs may not be the best one. And also, the process of discovering a drug that can block the activity of a protein such as the kappa-opioid receptor is a lengthy and expensive process. Computational tools can be employed to screen billions of compounds but it is a long-drawn process. Leslie Salas Estrada, from Icahn School of Medicine, proposes AI for Opioid treatment to optimize the process of blocking kappa-opioid receptors.

In her opinion, machine learning can process all the information available with large chemical databases and develop new drugs from scratch. This is apparently possible because ML algorithms can identify patterns from the big data and hence reduce the time and effort required. The algorithms extract information about the kappa-opioid receptor and the known drugs. The computer model is trained with a reinforcement learning algorithm that can identify properties favorable for drug treatment.

The team has identified several compounds having healing properties. They are planning to work with collaborators to synthesize them and test their ability to block receptors. When integrated with the healthcare system, the artificial intelligence tool can prove to be useful for reducing drug addiction, Salas hopes. Sometimes, present techniques can't identify the causes of risk there may be more involved in medical treatments than simply lowering a patient's opioid dosage.