A new study led by researchers with the Yale School of Public Health’s Department of Environmental Health Sciences and Imperial College London used network machine learning algorithms to identify chemical compounds in extra virgin olive oil (EVOO) that could aid in the prevention and treatment of Alzheimer’s disease.
The study identified 10 phytochemicals that appear to act similarly against Alzheimer’s disease protein networks as known pharmaceutical agents. One of the primary characteristics of Alzheimer’s disease is the abnormal build-up of proteins in and around brain cells. The EVOO molecules could potentially serve as a basis for future clinical studies, the researchers said.
The study was published in Human Genomics.
“Alzheimer's disease is a condition bearing an enormous human, social, and economic burden,” said Vasilis Vasiliou, the Susan Dwight Bliss Professor of Epidemiology and chair of the Yale School of Public Health’s Department of Environmental Health Sciences. “Our study, which integrates artificial intelligence (AI), analytical chemistry, and omics studies into a unique framework, provides fresh perspectives on how EVOO might contribute to the prevention and/or treatment of Alzheimer’s disease.”
Scientists with the National and Kapodistrian University of Athens collaborated on the research, which is part of a broader Yale-led effort to explore the health benefits of EVOO and other foods.
In the current study, the authors identified 67 bioactive chemicals in EVOO that could serve as potential candidates in preventing or treating Alzheimer’s disease. Using machine learning algorithms, they simulated how those chemical compounds might disrupt protein networks in the body that are related to Alzheimer’s disease. The researchers then identified a subset of chemicals that impacted the protein networks most similarly to existing clinically approved Alzheimer’s disease treatments, according to their machine-learning analysis.
While traditional machine learning has limitations in processing the network structures common in biological data, the network artificial intelligence used by the researchers in the current study was tailored to navigate these complexities. This adaptation could pave the way for more insightful analyses of disorders, including challenges like Alzheimer's disease, the study’s authors said.
Of the 10 phytochemicals listed in the study, quercetin had the highest likelihood of being as clinically active against Alzheimer’s disease as approved Alzheimer’s disease medications. Earlier studies have shown that this molecule/chemical decreases oxidative stress and inhibits the buildup of amyloid beta — the proteins linked with Alzheimer’s disease.
There are several limitations to the study. While the machine learning algorithm identified chemical compounds that appear to impact the cascade of protein networks caused by Alzheimer’s disease, the researchers were not able to gauge the overall effectiveness of their EVOO chemical compound candidates. Further research is needed to determine whether these chemicals are able to prevent or treat Alzheimer’s disease, the researchers said. And since the researchers only trained their algorithms on FDA-approved phase 3 or 4 drugs, there may be other potentially effective chemicals in EVOO that they may have missed.
Importantly, the researchers note that the integrated artificial intelligence, analytical chemistry, and omics analytical framework that was created as part of the study could serve as a valuable new tool in helping to advance the science surrounding the health benefits of EVOO.
“It is only through the conduct of such studies that the predictive utility of our machine learning approach will be validated,” the authors wrote in the study. “While the results of the present study shed light on how EVOO may help treat or prevent Alzheimer's disease, the same approach may be applied to identify EVOO phytochemicals (or other food constituents) that treat other diseases, such as dementia, hypertension or dyslipidemia.”
And the potential applications are not limited to EVOO, the researchers said.
“Recent advances in artificial intelligence, coupled with the explosive growth of large-scale, multi-source data on food, drugs, and diseases, offer a unique opportunity to identify molecules within foods that may prevent and/or reverse disease by using more holistic and systems-based approaches,” the study said.