AI-guided Drug Repurposing for Neurodegenerative and Neuroinflammatory Diseases

Abstract

Artificial Intelligence (AI), machine learning (ML), and deep learning (DL) have significantly revolutionized how drugs are repurposed. AI, ML, and DL aid in faster drug repurposing than traditional methods and lower costs by analyzing old and new drugs in a more data-driven approach than by random selection. Drug use through new therapeutic applications is beneficial for early disease detection of neurodegenerative (Alzheimer's and Parkinson's disease) and neuroinflammatory (multiple sclerosis, amyotrophic lateral sclerosis, etc.) diseases because of medical complexities and historical constraints with traditional drug development. Multi-omics datasets have allowed AI platforms to provide mechanisms for identifying novel drug-to-target-to-disease relationships. Platforms such as DeepDrug, COMIC, and DrugPipe use deep learning and knowledge graphs to create advanced connections among drugs, targets, and diseases that are not possible through traditional methods. By using a systems biology approach with combined information across multiple omics (genomics, transcriptomics, proteomics, metabolomics, etc.), a better understanding is gained of disease processes at a mechanistic level and the predictability of drug safety and efficacy in treating diseases. While drug repurposing is rapidly evolving, there remain significant challenges in clinical validation to identify disease targets and optimal drug candidates. The ability to make sense of complex model outputs (interpretability) and the need to establish standard protocols are essential for end-to-end drug discovery. AI-guided drug repurposing combined with systems biology is a new direction for drug discovery in neuropharmacology that may offer promising opportunities for enhancing drug discovery, improve patient outcomes, and accelerate medical breakthroughs for neurodegenerative, neuroinflammatory, and other debilitating conditions.