AI & Neurodegenerative Disease: Drug Repurposing
- Artificial intelligence may hold the key to identifying existing drugs that can be repurposed to treat Parkinson's disease, addressing the urgent need for effective therapeutics.
- Feixiong Cheng,PhD,head of the Cheng Laboratory and director of the Cleveland Clinic Genome Center,and his team used systems biology to cross-reference Parkinson's disease-related genetic variants with brain-specific DNA...
- The integrative AI model synthesized data from genetic, proteomic, pharmaceutical, and patient datasets to uncover hidden patterns.
AI Models Offer Hope in Repurposing Drugs for parkinson’s Disease
Updated June 10, 2025
Artificial intelligence may hold the key to identifying existing drugs that can be repurposed to treat Parkinson’s disease, addressing the urgent need for effective therapeutics. The lengthy clinical trial process for new Parkinson’s disease treatments frequently enough takes over a decade.
AI models are being used to identify existing drugs that could be repurposed for Parkinson’s disease treatment.
Feixiong Cheng,PhD,head of the Cheng Laboratory and director of the Cleveland Clinic Genome Center,and his team used systems biology to cross-reference Parkinson’s disease-related genetic variants with brain-specific DNA and gene expression databases. They then combined these findings with protein datasets to assess which identified genes affect proteins in the brain, pinpointing several known to cause inflammation when mutated.
The integrative AI model synthesized data from genetic, proteomic, pharmaceutical, and patient datasets to uncover hidden patterns. This process generated a list of potential candidate drugs. Subsequent analysis of electronic health records revealed that individuals prescribed the cholesterol-lowering drug simvastatin were less likely to recieve a Parkinson’s disease diagnosis.
The findings were published in NPJ Parkinson’s Disease earlier this year.
We lack effective treatments for PD, and ther are no good experimental models for PD drug discovery. AI-based computational methods offer effective strategies for PD target and drug discovery.
Feixiong Cheng, PhD, Cleveland Clinic Genome Center
Cheng noted that conventional genome-wide association studies have identified many Parkinson’s disease risk variants in non-coding human genomes. He added that AI models can integrate functional genomics data to identify likely causal genes for Parkinson’s disease as potential drug targets.
According to Cheng, there is a high demand for repurposing drugs for conditions like ALS and Alzheimer’s disease. He noted that, based on their 2025 Alzheimer’s disease pipeline, 33% of compounds under clinical trials for Alzheimer’s are repurposed drugs.
Cheng believes AI models could be trained to identify clinical failure drugs with good safety profiles for other diseases.He also mentioned that AI can design novel small molecules or antibodies for challenging chronic diseases like Alzheimer’s, Parkinson’s, and cancer.
What’s next
Cheng’s team is exploring the use of generative AI to design novel antibodies as effective biological therapeutics for Alzheimer’s disease, expanding the potential applications of AI in drug discovery and Parkinson’s disease treatment.