Artificial intelligence (AI) and whole genome analysis are poised to turn the world of healthcare on its head. Advances in recent years have taken the cutting-edge technologies from science to reality, and one company is now using them to dramatically improve patient outcomes and make personalized medicine the norm. Already working with doctors and hospitals to mine DNA and use AI to provide better treatments much faster, Frélii, is working on its next goal: building an integration tool for medical devices so that the information generated from a device can be analyzed and delivered to the doctor in a more streamlined fashion.
“Frélii’s technology has been designed to allow assessment and sequencing that could quickly change personalized medicine,” says Ian Jenkins, CEO of the company. Frélii is essentially data mining DNA. The company leverages the big data of DNA to improve health and wellness and enhance the human experience.”
According to Jenkins, Frélii uses its unique technology to take a fundamentally different approach than what others in the market are doing. It leverages its AI to interpret DNA testing results, delivering much greater insight and decision-making perspective than what is currently available. The company enables visualization into epigenetic expression as clusters of millions of 3D models. Through this, its AI derives meaning from the clusters with true interconnected data. As such, Frélii opens the door to far more specific and accurate DNA analysis and far broader applications to a number of market segments.
“We aren’t the first to select based on genetics, but we are the first to be able to isolate and determine why a test is not succeeding,” says Jenkins. “We can bridge the gap of AI genetics to see how genetics are expressing or changing the interaction of a drug. That is a very valuable component to these people because right now, it’s kind of a crap shoot. If you think about clinical trials, they might want someone with rheumatoid arthritis in a certain age range. The problem is that everyone is unique and will respond differently. If you look at traditional pharmacogenetic tests, they are really basic. You may have one mutation and you need to look at what that mutation means.”
To bring this into perspective, current competitive solutions analyze from 400,000 to 750,000 data points, generating approximately 384 outcomes, such as diet suggestions or identification of potential health risks. The Frélii AI technology has dramatically increased the data points analyzed to about 3.2 billlion; yielding exponential numbers of potential outcomes than what has been available to doctors, researchers, and patients. According to the company, this increase represents a paradigm-changing jump in the availability of accurate, whole-genome sequencing and analysis.
Frélii’s Quantum Leap
Jenkins says AI looks at each mutation in relation to every other genotype and mutation in the genome because they are connected in some way or another. “You have to value those mutations based on each other. For instance, take BRCA, a common expression in a cancer patient. Right now, a lot of people are looking at maybe seven base pairs. There are 2000 known mutations, and they all interact to make one more important or less important, and you don’t know how important ‘that one’ is until you have the other one,” he says. “The key is looking at every genotype and the connections between them. And that is the quantum leap that Frélii is bringing to the table,” says Jenkins.
Jenkins says that before Frélii, no one could never quantify how the genomes interacted with each other. Moreover, he says, there was a lot of data out there, but it was too complicated, cost prohibitive, and took too long to analyze.
“When we first started doing this, it took about 6–8 weeks to analyze a genome, then about 166 hours to run the AI. With AI, it was much cheaper than doing it manually, but it was still difficult.”
Frélii’s new type of compression algorithm that now runs in two hours — what Jenkins call “a very significant compression.” Over the last few years, AI has significantly improved, and it is much more powerful. It can sequence a different type of methodology in 5–7 days and run the analysis in 2 hours. Frélii’s predictive capacity with its AI has increased from 84 to 98.5 percent. And its high-efficiency genetic sequencing and analysis using Frélii’s proprietary technology has increased to greater than 99 percent and 99.999 percent accuracy on whole genome and exome sequencing, respectively.
According to the company, Frélii conducts whole genome sequencing of 3.2 billion individual data points. After sequencing, Frélii runs the data through its patented artificial intelligence “Navii.” Navii’s complete analysis includes analyzing every data point against all existing published genomic studies, as well as analyzing every data point against all other data points. This process takes more than 7,680,000,000 gigs of computational processing per individual and delivers the most comprehensive genetic results on the planet. It helps Frélii provide doctors real answers to potential patient needs or risks.
The data is accessible to all of the doctors and hospitals in a group, providing a more seamless integration of the data. The main tool analyzes about 27,000 diseases, and it allows doctors to search, looking at interaction and even genome. That’s where precision medicine benefits: the doctors can pinpoint the dose that is just right for the patient.
“What AI is doing is really helping doctors be more specific.” He says that AI has been especially beneficial in the assisting doctors who are prescribing cannabis for late stage cancers. He says it is also particularly helpful in guiding doctors who are prescribing psychotropics.
“With psychotropics, the doctor will typically try something and have the patient come back in a month. It is common to take 6-9 months to get the right combination. But, if you’re able to analyze the data much faster, you’ll be 99 percent the way there in the first two weeks. It’s a drastic time reduction for the patient. And once they are in the database, if something else happens, there is no lag time. During a visit, the doctor, can go data mine through his own portal and make recommendations the patient.
Integrating Medical Device Data
The next thing for Frélii is a tool that integrates medical device data into the AI. “The AI doesn’t just have to take in DNA. We tested it on taking in blood and medical device data. Just like DNA, it can take that information and read that in and analyze the outcome the doctor is seeing and make suggestions,” says Jenkins.
The goal, he says, is that AI will ultimately integrate with every medical device. “We want all medical devices integrated into one unified structure and doctors can use the AI, essentially what it will do is improve the delivery of the medical device data to the doctor by linking it to our AI,” says Jenkins. “AI can read it in and use that big data over time to give the doctor a better view of the patient. When they see everything in a concise manner, it makes that device much more valuable and makes the doctor’s diagnosis that much faster.”
Jenkins says Frélii has begun testing with medical devices but won’t be releasing the full integration product for at least 12 months. The system would link the medical device data back to the DNA and to the doctors’ data so that all of them could be analyzed simultaneously.
“Give us a couple of years, and the medical industry is going to change dramatically,” says Jenkins.
This article was written by Sherrie Trigg, Editor and Director of Medical Content for MDB.