Cleveland Clinic's Genome Centre, in conjunction with IBM, is looking into new ways to treat chronic pain, which affects roughly one in every five Americans. The initiative, led by Feixiong Cheng, PhD, and a team of academics, seeks to develop non-addictive pain therapies through the use of artificial intelligence. The findings, published in Cell Press, emphasise the potential of artificial intelligence in repurposing existing FDA-approved medications for enhanced pain management.

The study uses deep-learning algorithms to discover gut microbiome-derived compounds and already licensed medications that can interact with pain receptors, providing an alternative to opioids. Chronic pain therapy remains difficult due to the hazards associated with opioids, such as serious side effects and addiction. Co-first author Yunguang Qiu, PhD, a postdoctoral fellow in Dr. Cheng's group, stated, "Recent evidence has shown that targeting a subset of pain receptors in G protein-coupled receptors (GPCRs) can provide non-addictive relief." The challenge is to determine how to target those receptors.

 Instead of creating novel compounds, the researchers concentrated on repurposing current medications. Part of this technique involved mapping gut metabolites to identify medication targets. Dr. Qiu and computational scientist Yuxin Yang, PhD, led a team that updated a previous AI algorithm at the Cheng Lab to achieve this. Dr. Yang stated, "Our IBM associates gave us excellent advice on developing sophisticated computational techniques. I appreciate the opportunity to learn from peers in the industry field."

 The study team developed the LISA-CPI technology, which uses artificial intelligence to predict how chemicals attach to pain receptors. As Dr. Cheng put it, "AI can rapidly analyse compound and protein data from imaging, evolutionary, and chemical experiments to predict which compound has the best chance of influencing pain receptors in the right way."

 The scientists evaluated 369 gut microbial metabolites and 2,308 FDA-approved medicines using LISA-CPI, identifying numerous molecules with potential for pain treatment. The technology lightens researchers' experimental burden, allowing them to efficiently examine more medications and substances. Dr. Yang pointed out: "This algorithm's predictions can lessen the experimental burden researchers face when developing a list of candidate drugs for further testing."

 Looking ahead, Dr. Cheng emphasised AI's larger potential in drug discovery, saying, "We believe these foundation models will offer powerful AI technologies to rapidly develop therapeutics for multiple challenging human health issues."