One of medicine’s most perplexing problems is Alzheimer’s disease, a neurodegenerative disorder that causes dementia, memory loss, personality changes, and other irreversible symptoms. Medicines can treat the symptoms of the disease, but a cure has been a challenge – perhaps because the cause of Alzheimer’s disease is unclear.
leading group Rui Chang, Ph.D.; An associate professor at the University of Arizona College of Medicine – Tucson, along with collaborators at Harvard University, have used artificial intelligence to determine the cause of Alzheimer’s disease, inferring the molecular changes that healthy neurons experience in the human brain as the disease progresses. The results are published today in Nature Communications Biology.
A road map to track the journey of Alzheimer’s
“There are several pathways involved in Alzheimer’s disease,” said Dr. Chang, referring to the sequence of events that occur in cells to trigger changes in the body. Previous studies have shown a link between these brain abnormalities and the disease, so for a long time researchers have focused on the formation of abnormal structures in the “amyloid plaques” and “tau tangles” that form in the Alzheimer’s brain. However, drugs that clear and stop the production of plaques and tangles have failed in clinical trials – suggesting that they are not the cause of Alzheimer’s but a consequence of earlier events.
Dr. Chang compares the path from health to Alzheimer’s, where plaques and tangles respond to problems “upstream,” likening the flow of water “downstream.”
“Amyloid plaques and tau tangles are immediate effects of the series of genetic mutations that produce Alzheimer’s disease. It is highly doubtful that direct targeting of these abnormal structures will be effective. In my view, the right way is to target the disease upstream,” said Dr. Chang. “So it’s important to understand the whole picture.”
Dr. Chang uses AI to design this landscape. With tissue samples taken from more than 2,000 Alzheimer’s brains from a national database, Dr. Chang’s AI algorithm retrieved a mathematical network model of the human brain, extracting in-depth information about genetic and molecular processes. His team can now map genes that work together across the entire genome and track changes in the interactions of these genes as Alzheimer’s progresses, providing clues to the origins of the disease and tracing the molecular pathway from health to disease.
“AI is a novel method that teases out vast amounts of data into a network model that reveals which upstream genes regulate important downstream genes to provide a clear picture of upstream events,” said Dr. Chang. “With that model, we can identify the upstream genes that trigger amyloid plaques and the downstream tangles. These upstream genes may be better targets for potential therapies.”
Examining these genes individually in a traditional laboratory is time-consuming, so the scientists’ approach is to use high-powered computing and AI to select the most promising targets for precise treatment.
“This is not studying one gene at a time – it’s 6,000 targets at once, which greatly accelerates drug development and discovery,” said Dr. Chang. “This is the first study to show that AI and a big data-based approach can open the door to developing new ways or pathways to develop Alzheimer’s treatments.”
Finding targets and drugs that affect them
Dr. Chang used AI to identify 19 neuron-specific genetic points of interest in the Alzheimer’s pathway, which predispose neurons to the disease state. Research collaborators at Harvard confirmed the role of these genes in the development of Alzheimer’s by using stem cells to create neurons in a petri dish and then turning off the genes to see what would happen. They found that 10 of these genes affect the production of plaques and tangles and could be explored as targets for treating Alzheimer’s disease.
“If inactivating these genes significantly alters the amount of amyloid plaques and tau tangles, this would be a promising target for developing Alzheimer’s disease therapies,” Dr. Chang said.
Once the gene targets are identified, the next step is to find drugs that attack those targets. Again, high computational power is used instead of tedious and time-consuming laboratory experiments. 3D computer models determine whether existing molecules and drugs fit into drug targets like a key-to-key lock.
These experiments helped the team screen millions of Food and Drug Administration-approved, natural product, and small-molecule compounds against more than 6,000 targets, narrowing the field to about 3,000 candidates of interest. They are investigating further with several small molecules, and the team has a National Institutes of Health grant that will enable clinical trials on three of the compounds. They expect human testing to begin soon.
“From mathematics and data, I can design mathematical algorithms to take large amounts of data from patients to clinical studies,” said Dr. Chang. “It’s been an exciting journey for me to see compounds move into clinical trials and ultimately benefit patients.”
Using computing power to solve health problems
The AI was originally coded by Dr. Chang and used to conduct similar studies in conditions such as melanoma, diabetes and cardiovascular disease, making his team the first to use artificial intelligence to find genetic targets in various diseases. He is cautiously optimistic that AI can speed up precision drug research in many of medicine’s vexing problems.
“We hope AI is a game changer,” said Dr. Chang. “I hope that within five years there will be more clinical trials to treat Alzheimer’s disease and within 10 years there will be two drugs approved by the FDA to stop or reverse disease progression.”
Dr. Chang specialized in Computer Science, AI and Machine Learning with his PhD, and has been applying his skills to genomic data for over 15 years. He said the opportunity to work at the intersection of computing and biology is extremely rewarding.
“The mind is an amazing organ because it defines us as human beings,” he said. There are many complex functions yet to be discovered that can open some hidden doors.
This study was conducted using a database provided by the Accelerating Medicines Partnership® Program for Alzheimer’s Disease (AMP® AD), a public-private partnership between the National Institutes of Health (NIH), the US Food and Drug Administration (FDA). A number of biopharmaceutical and life science companies and non-profit organizations are working to change the current model to develop new diagnostics and treatments for Alzheimer’s disease. This work was supported in part by the National Institute on Aging, Division of the National Institutes of Health, under Award Nos. 1R56AG062620-01, 1RF1AG057457-01, R01AG055909, and RF1NS117446 from the National Division of Neurological Disorders and Stroke. NIH, under award number U54NS110435, and UArizona for Innovation in Brain Science.
[ad_2]