Researchers use AI to find non-opioid pain relief options

It is estimated that one in five Americans suffers from chronic pain, and current treatments leave many gaps. Dr. Feixiong Chen, director of the Cleveland Clinic Genome Center, and IBM are leveraging artificial intelligence (AI) for drug discovery in advanced pain management. The team’s deep learning framework identified multiple gut microbiome-derived metabolites and FDA-approved drugs that can be repurposed to select non-addictive, non-opioid options to treat chronic pain.

The findings, published in Cell Press, represent one of the many ways the organization’s Discovery Accelerator partnership is contributing to advances in healthcare and life sciences research.

Treating chronic pain with opioids remains a challenge because of the risk of severe side effects and dependence, said co-lead author Dr. Yunguang Qiu. He is a postdoctoral fellow in Dr. Chen’s lab, and his research program focuses on developing treatments for neurological diseases. System failure. Recent evidence has shown that drug administration of a specific subset of pain receptors from a class of proteins called G protein-coupled receptors (GPCRs) produces non-dependent, non-opioid analgesic effects. The question is how to target these receptors, Dr. Qiu explains.

Rather than inventing new molecules from scratch, the researchers wondered if they could apply the research methods they had already developed to find existing FDA-approved drugs for potential pain indications. Part of this process involves mapping gut metabolites to identify drug targets.

To identify these molecules, lead author Dr. Yuxin Yang, a computational scientist and former Kent State University graduate student, conducted research. Dr. Yang completed his dissertation research in Dr. Chen’s lab and continues to work there as a data scientist. Doctors. Yang and Qiu led their team to update a previous drug discovery AI algorithm developed by Cheng Lab. Collaborators at IBM helped write and edit the manuscript.

“Our collaborators at IBM provided valuable advice and perspective to develop advanced computational techniques,” says Dr. Yang. “I am pleased to have the opportunity to collaborate and learn from colleagues in the industry.”

To determine whether a molecule acts as a drug, researchers need to determine how it physically interacts with and affects proteins in the body, in this case pain receptors. Must be predicted. To do this, researchers need to understand both molecules in 3D, based on extensive 2D data about their physical, structural, and chemical properties.

“Even with current computational techniques, combining the amount of data required for predictive analysis is extremely complex and time-consuming,” explains Dr. Chen. “AI can rapidly leverage both compound and protein data from imaging, evolution, and chemistry experiments to determine which compounds are most likely to have the appropriate impact on our pain receptors. can be predicted.”

The research team’s tool is called LISA-CPI (Ligand Image and Receptor Three-Dimensional (3D) Structural Recognition Framework for Predicting Compound-Protein Interactions), a form of artificial intelligence called deep learning. Use to predict:

Whether a molecule can bind to a particular pain receptor Where on the receptor the molecule physically binds How tightly the molecule binds to that receptor When a molecule binds to a receptor, the signaling effect is turned on or turn off

The research team used LISA-CPI to predict how 369 gut microbial metabolites and 2,308 FDA-approved drugs interact with 13 pain-related receptors. The AI ​​framework has identified several compounds that can be repurposed for pain treatment. Research is underway to test these compounds in the laboratory.

“The algorithm’s predictions can reduce the experimental burden that researchers must overcome to generate a list of drug candidates for further testing,” says Dr. Yang. “We can use this tool to test even more drugs, metabolites, GPCRs, and other receptors to find treatments that treat diseases beyond pain, like Alzheimer’s disease.”

Dr. Cheng added that this is just one example of how the team is working with IBM to develop small molecule fundamental models for drug development. This includes both drug repurposing in this study and ongoing novel drug discovery projects.

“We believe these foundational models provide powerful AI technologies to rapidly develop treatments for multiple difficult human health problems,” he says.