The effort is led by Feixiong Cheng, PhD, director of the Genome Center at Cleveland Clinic. Through advanced computer-based systems and analytical tools, Dr. Cheng and his research team analyze databases of human gene sequences and molecular targets to uncover potential new uses for established medications. The most promising AI-predicted repurposable drugs are then tested using large-scale data.
The research team’s most recent paper, published in npj Parkinson’s Disease, highlights this approach, revealing that a common cholesterol drug may have potential benefits for Parkinson’s disease.
In the study, Dr. Cheng’s team used AI models to integrate genetic, proteomic, pharmaceutical and patient datasets, identifying patterns that may not be apparent when examining one type of data alone. By cross-referencing their findings with pharmaceutical databases and EHRs, they discovered multiple candidate drugs and assessed differences in Parkinson’s diagnoses among patients taking these medications.
According to Dr. Cheng, this method enables researchers to uncover new drug candidates at an unprecedented scale and speed.
“The timing is crucial—we can conduct this kind of research in just a few months and identify potential drugs for further testing,” Dr. Cheng told Becker’s. “The scalability is also significant. Our AI approach isn’t limited to a single drug or disease; we can screen all FDA-approved drugs—around 2,000 of them—not just for Parkinson’s but for many other conditions.”
One of the most promising drug candidates identified by the Cleveland Clinic team was simvastatin, a widely used statin for lowering cholesterol.
“We found that individuals taking simvastatin had a significantly lower risk of developing Parkinson’s disease,” Dr. Cheng said. “Our next step is to conduct further clinical testing to verify its potential efficacy in early-stage Parkinson’s patients.”
During the COVID-19 pandemic, Cleveland Clinic researchers used this method to identify a sleep disorder medication, modafinil, as a potential treatment for COVID-19. Their AI models also suggested sildenafil, widely known for treating erectile dysfunction, as a candidate for Alzheimer’s disease treatment—findings that have been validated in both patient cell models and animal studies.
“We are preparing for a Phase 2 clinical trial to further test sildenafil in Alzheimer’s patients at Cleveland Clinic,” Dr. Cheng said. “While Parkinson’s research is relatively new for us, we are applying the same methodology to identify promising candidates.”
However, funding for clinical trials remains a significant hurdle. Since repurposed drugs lack patent protection, pharmaceutical companies often do not support such studies.
“This is great for patients because repurposed drugs are much cheaper, but it also means we rely on funding from government agencies, foundations or philanthropic sources to push these studies forward,” Dr. Cheng said.
Dr. Cheng’s team continues to explore AI’s potential in drug repurposing for neurodegenerative diseases. Their findings suggest that reducing cholesterol, glucose levels and blood pressure through existing medications could lower the risk of Parkinson’s and other conditions.
“Parkinson’s is a complex disease with multiple risk factors,” Dr. Cheng said. “We are still investigating the exact mechanisms, but our AI-driven approach provides a powerful tool to accelerate the discovery of effective treatments.”