AI Uncovers a Key Gene in Alzheimer’s Disease and a Potential Treatment Researchers at the University of California, San Diego (UCSD) have made a significant discovery regarding a gene that plays a direct role in causing Alzheimer's disease, a degenerative brain disorder affecting over 7 million Americans. Their findings, published in April in the scientific journal Cell, suggest that the gene PHGDH, previously thought to be just a biomarker, is in fact a causal factor in "spontaneous" Alzheimer’s, which accounts for most cases of the disease. This breakthrough could pave the way for new and more effective treatments. While certain gene mutations are known to increase the risk of Alzheimer’s, these genetic factors explain only a small fraction of the total cases. Therefore, understanding the spontaneous causes of the disease is crucial for developing better medical interventions. According to senior author Sheng Zhong, a professor in the Shu Chien-Gene Lay Department of Bioengineering at the UCSD Jacobs School of Engineering, current treatment options for Alzheimer’s are limited, and their effectiveness is often poor. To uncover the role of PHGDH, the UCSD team utilized advanced artificial intelligence (AI) technologies. They employed powerful computational models, such as the Latent Prompt Transformer, on high-performance NVIDIA RTX A6000 and H100 GPUs. These tools enabled them to analyze complex biological data and understand how PHGDH affects gene regulation in the brain, leading to the onset and progression of Alzheimer’s. One of the key insights from the research is that PHGDH can trigger harmful cellular changes associated with Alzheimer’s. By disrupting the brain’s normal gene regulatory processes, the gene contributes to the disease’s development. The AI models provided a detailed view of these interactions, highlighting PHGDH’s dual role as both a marker and a causative agent. Moreover, the AI-driven approach identified a promising therapeutic candidate, a small molecule called NCT-503. This molecule can inhibit PHGDH’s damaging effects without disrupting its essential functions. Tests conducted on mice showed that NCT-503 significantly slowed the progression of Alzheimer’s and improved cognitive functions. These results underscore the transformative potential of AI in medical research, particularly in unraveling complex biological mechanisms that were previously challenging to study. "This breakthrough offers a new therapeutic candidate with demonstrated efficacy," Zhong noted. "There may also be entirely new classes of small molecules that can be further developed for future therapeutics." The researchers are now focused on optimizing NCT-503 and conducting additional pre-clinical studies before seeking FDA approval for a potential new drug. The identification of PHGDH’s role in Alzheimer’s and the discovery of a potential treatment highlight the increasing importance of AI in advancing medical science. These technologies are not only aiding in understanding complex diseases but are also accelerating the development of targeted therapies, offering hope for millions affected by Alzheimer’s disease. For more details, you can read the full research paper in Cell or explore additional coverage of the UCSD research.