Unlocking the power of AI: Targeting cellular health for anti-aging products

01 Sep 2023 --- Following the increasing use of AI to develop products faster, with a higher efficacy, biotechnology company Science Research Wellness (SRW) is using the technology to create products that target people’s cellular age.

In this first of a two-part series, Nutrition Insight sits down with Greg Macpherson, SRW’s founder, to discuss how AI can help find anti-aging pathways and screen molecules that modulate these processes.

“We started off using AI for regular intelligence, using it to validate compounds that we’re using, to determine that they’re hitting the pathways that we anticipate and also whether they’re stimulating the downstream effects of what we’re hoping to achieve.”

“We also work in the DNA methylation space. We have a partner that uses its AI machine learning to validate the DNA methylation patterns associated with aging.”

Macpherson explains that this refers to humans’ biological aging clock; the patterns on a person’s DNA reflect that clock.

In addition, SRW partnered with an AI drug discovery company to help develop products much faster, with a higher level of clinical effect; some companies “can do ten years of research in three months.”

“The race is on”
To succeed in the next decade, companies must focus on AI and machine learning, stresses Macpherson, “because that’s where the breakthroughs are going to come from, especially in this particular field.”

Greg Macpherson, founder of SRW (Image credit: SRW).“We’re all looking and seeking to modulate aging because we know that aging delivers the biggest risk of health compared to health conditions.”

He underscores that the fastest way to understand the compounds, especially in nature, that interact with the pathways associated with aging is to work with AI. “In those AI or machine learning systems, we can plug in what pathways we want to interact with.”

“The race is on to identify nutrients or groups of nutrients, which can benefit the aging process,” Macpherson continues. “Then we have to lean into the DNA methylation patterns and the effect that those nutrients or nutraceuticals have on those on the DNA methylation. This gives us clues and signals that our formulations are working effectively.”

“Probably ten years ago, if someone said I was going to do an anti-aging intervention, we’d likely have to spend 50 years putting people on it to pick up enough to determine whether it works. With DNA methylation, we can very quickly work out if we change the methylation patterns toward someone biologically younger.”

He details that through this process, researchers can do trials that would have taken 50 to 60 years in the past in six to twelve months. “This can massively accelerate how quickly we can identify what’s working and not.”

Biological aging clocks
SRW also works with DNA methylation tests to determine someone’s biological age. If cells are aging, it is more likely someone will encounter a disease associated with aging, explains Macpherson.

“I can take a blood cell of someone and run it through our machines to give a biological age, which is more reflective of actually how old their cells are acting. That can be quite different from your chronological age. Someone could be 40 but have a biological age of 60 or 30.”

“You can only do that through machine learning,” Macpherson underscores. “Because of the size of the data sets and complexity of DNA, you just could not do it without having AI deeply embedded in that process.”

In addition, different biomarkers also leave patterns in a person’s DNA methylation.

“What’s happening next is that we can check your DNA methylation status and not only get a biological age, but we can also start to get tissue ages, such as how old your brain is acting. How old is your heart or liver, etc.”

“With that level of granularity, we can start to understand what compounds impact a heart or a brain aging. That gets super exciting because we can then essentially find out what the status is for all of the tissues in your body.”

Research published earlier this year demonstrated that clinical trial participants decreased their biological age by 4.6 years on average in an eight-week lifestyle and diet intervention program.

To succeed in the next decade, companies must focus on AI and machine learning, stresses Macpherson.Preventative care
Macpherson notes that diseases do not happen overnight but occur over 20 years. “If we can pick up a signal that perhaps someone’s heart is aging a bit faster than the rest of their body, we can stage an intervention with molecules that AI has picked up that impact that organ age. And then perhaps we can actually change the process.”

He suggests that it may be possible to shift the progression of someone’s disease so they do not have to experience it.

“There’s a lot of work to be done to make that happen. But this is the opportunity that AI will bring that will revolutionize what health means, and hopefully, we can get to preventative health care.”

Macpherson further states that people who work with AI are characterizing proteins, determining how they work and then finding the molecular structures of the compounds they’re screening to determine how compounds interact with those proteins.

They are discovering molecules “all the time,” he notes, which unexpectedly interact with those receptors or proteins.

For example, AI peptide specialist Nuritas developed a whey protein that uses its plant peptide to enhance the protein used in sports nutrition.

Road to commercialization
AI can help to screen compounds with potential use in nutraceuticals, after which it needs to be validated with additional testing, explains Macpherson.

“You might screen 10,000 compounds and get hits on ten of those. Then, you should research it in a petri dish to understand what’s happening at a cellular level. Perhaps then you’re starting to look at it from the Caenorhabditis elegans worm and move into mice, etc. At each step, you’re validating what’s got the best hits.”

AI can help screen compounds for nutraceuticals, after which it needs to be validated in additional testing.However, he cautions that machine learning is only as good as the algorithm driving it. “This is why you might have ten hits, but you need to validate those to ensure it’s working, learn from it and train the AI.”

“What’s interesting is that something like 40-50% of compounds that have some anti-aging effect, even in tests on worms, will translate into some health benefit, at least in mice. Then you have to link it to humans.”

The knowledge needs to be translated into a product people can benefit from. Researchers can repeat the process and refine a product’s formula through machine learning.

Accelerating knowledge
AI can help to accelerate what we know, highlights Macpherson. “Many of the molecules that we know and love that are good for us, such as curcumin, impact various anti-aging pathways.”

He suggests curcumin, which has been touted for its anti-inflammatory properties, may drive some anti-aging processes. Moreover, the wide use of curcumin in food and nutrition demonstrates that it is safe for human consumption.

“That was the logic for how we designed our products initially. We looked at what compounds were interacting with pathways associated with aging.”

He concludes: “We could bring them to market quickly because we knew they had been used for a long time. It was just a combination we put them in that was unique.”

By Jolanda van Hal

To contact our editorial team please email us at editorial@cnsmedia.com