Imagine you’re driving in busy downtown traffic. Traffic lights are switching from red to green, keeping the cars moving.

Now picture the inside of your brain working in a similar way, except there are billions of these signals switching on and off like traffic lights to keep your thoughts flowing smoothly.

In diseases like Alzheimer’s, some of these signals can fail or switch out of sync, creating a traffic jam in the brain. According to the Alzheimer's Association, almost 7 million Americans are living with this devastating disease. At the University of Toledo College of Medicine and Life Sciences, we are using artificial intelligence, now commonly called AI, to help decode brain signals and restore the brain’s health.

Your brain cells communicate through complex roads of molecular proteins called signaling networks. In these networks, special proteins called kinases are like traffic controllers, turning traffic signals on or off. These kinases signal other proteins on the road when to go and when to stop, sending the right messages down the road along with other protein messengers. This is why your brain cells know what to do and when to do it.

We have learned that Alzheimer’s, schizophrenia, and clinical depression are linked to misfiring of kinase signals. Fixing those misfiring signals could one day help treat these brain disorders.

But to reach this goal, we first need to map out what each kinase signaling road looks like normally, so when things do go wrong, we know exactly what signals we need to fix.

How is it possible to do this kinase mapping when each of your brain cells have thousands of these kinase signals active at any one moment, creating a mountain of biological data? It’s like trying to solve a jigsaw puzzle with many thousands of clear pieces and no picture on the puzzle box to help you! Examining just one kinase signaling pathway at a time is going to be slow work.

Fortunately, this is where AI becomes very useful.

My research field is called bioinformatics and is where data science meets biology. AI is a powerful tool that we bioinformaticists use to turn large amounts of kinase signaling data into useful results for our research into brain disorders linked with kinase signaling disorders.

For example, I have recently developed a new bioinformatics tool with specialized AI called PAVER and used it map kinase signaling in two types of brain cells: astrocytes and microglia.

These cell types may not get as much attention as nerve cells in the brain but are just as important to your brain’s health.

Astrocytes help support nerve cells by maintaining the blood-brain barrier, the shield around your brain that protects it from harmful things in the blood. Microglia are an important part of the brain’s immune system, cleaning up waste like damaged cells and responding to infections that sneak through the blood-brain barrier. If anything goes wrong with the function of astrocytes or microglia, inflammation can occur in the brain, which can lead to different diseases like Alzheimer’s.

We studied astrocytes and microglia by measuring dozens of kinase activities within these specialized brain cells at the same time. Then, we used my newly developed bioinformatics AI tool to map out their unique signaling networks.

The AI-generated results of my biological studies are still ongoing, but we have already discovered that microglia and astrocytes each possess unique kinase signaling maps, like having their own special traffic network controlling cellular communication. For example, microglia are actively involved in signaling pathways for releasing nerve cell growth factors and recruiting other immune cells to help with infection, whereas astrocytes are actively involved in pathways for nerve cell communication and inflammatory responses.

By using AI to detect which kinases were actively directing traffic, we uncovered critical signals that traditional methods would have missed. Interestingly, certain kinases could be detected by genetic methods, but our AI methods were also able to detect if they are active or if they are inactive, like traffic lights that exist but never turn green or red.

These newly identified active signals discovered by AI will likely help to open the door to personalized medicine in the future. If we can map specific abnormal signaling patterns in a patient’s brain, doctors could then tailor treatments to their specific needs. You can think of it as finding which traffic lights are out before sending the repair crew. By using AI to pinpoint specific molecular traffic jams, a therapy can be designed to clear them up more effectively and with fewer side effects.

At the end of the day, AI is helping us put this puzzle together correctly. Every new piece we fit into place brings us one step closer to understanding how the brain works, and how to heal it when something goes wrong.

William (Billy) George Ryan V is a Ph.D. candidate in the University of Toledo college of medicine and life sciences biomedical science program’s bioinformatics track. Billy is doing his Ph.D. research in the lab of Dr. Robert Smith in the department of neurosciences and psychiatry. For more information contact Billy at wryan3@rockets.utoledo.edu or go to utoledo.edu/med/grad.

First Published April 7, 2025, 3:30 a.m.