Pinpointing ingredients’ impact: US$43m AI platform to slash R&D process “by years”
24 Jul 2020 --- MedicascyAI is a new artificial intelligence (AI) solutions platform that works within a matter of weeks to discover where ingredients can have a significant impact. efficacyAI, the company that has the exclusive license, says that this will cut R&D time in the nutraceutical, supplement and cannabis industries by months or even years, while also improving the product. The platform has been in the making for over 13 years and is backed by US$43 million from the US National Institutes of Health (NIH), Georgia Tech, private donors and the Georgia Research Alliance.
“When MedicasyAI is applied at the beginning of the process to a library of molecules or ingredients, it can identify those that have the highest confidence predictions for being safe and effective. In turn, the manufacturer or nutraceutical company can focus its efforts on the ingredients that can give the best potential outcome. The company’s resources can then be applied to testing those few with the highest confidence prediction, saving time and money,” Tony Bellezza, Chairman and CEO of efficacyAI, tells NutritionInsight.
He adds that efficacyAI is now working with several prominent nutraceutical and cannabis companies on products that have the potential for a significant impact and benefit to consumers. There is currently a business plan for efficacyAI to commercialize the use of MedicascyAI in all regions. This growth strategy involves first working with US-based companies before building connections in international areas.
The technology works by screening the interactions of molecules with every human protein.“There will always be competition in this technology space of AI. However, unlike others, MedicascyAI is a technology solutions platform that only requires the chemical structure of a molecule. With this information, it then provides high confidence predictions on safety and effectiveness,” Jeffrey Skolnick, Creator of MedicascyAI, also tells NutritionInsight.
He adds that MedicascyAI is much faster and outperforms the competition. “While it may not cover the whole spectrum of natural ingredients, we can tell you with high confidence which items in a library will go after a specific benefit. The competition requires far more experimental information before making safety and effectiveness predictions. After you’ve invested all that time and money in the preliminary research, their algorithms can only predict yes or no.”
Steering toward more successful results
According to the company, MedicascyAI will reduce the nutraceutical industry’s reliance on secondary science and guide primary research toward more successful results. The technology works by screening the interactions of molecules with every human protein. MedicascyAI’s results are documented in over 50 major peer-reviewed papers.
“A given molecule can interact with 58 different protein families. Some may do nothing. Some may be antagonistic and some may be helpful. We look at how those interactions may translate into potential benefits and side effects, as well as mode-of-action targets,” explains Skolnick.
The tool also has confidence indices (CIs), which predict the probability that the prediction is correct. Skolnick estimates that MedicascyAI’s algorithms cover about 97 percent of all proteins and make reliable predictions for about 40 percent of molecules found in natural products.
According to Skolnick, MedicascyAI can:
- Predict what a molecule is good for. “It tells us when we may be right, which is about 70 percent of the time, and when we can ignore the predictions.”
- Suggest testing known supplements for other health benefits. “We have a high probability of accuracy in about 40 percent of these cases. With a minor investment, you may be able to repurpose your product for something you may not have considered.”
- Identify side effects. “We are correct about 78 percent of the time in identifying the worst side effects. While we can’t say what percentage of people will suffer the worst side effects, this tool will help you assess if the risk is appropriate to the benefit you’re trying to achieve.”
MedicascyAI is employed at the start of the R&D process. When used on a library of molecules or ingredients, MedicascyAI provides various targets, with companies honing in on those with the highest confidence predictions for safety and effectiveness. “A company wants to focus only on those having the highest confidence predictions. Research and testing in animals or humans is still the best course to test out the ingredient,” Skolnick further adds.
Overcoming challenges
There is significant latent demand for non-drug solutions.
Bellezza explains that, as is often the case with other start-ups or emerging and disruptive technology, the main challenge is helping people understand how MedicascyAI impacts their business model or industry.
“To build trust, we have done several proof-of-concept projects that clearly demonstrate how we can add value to a nutraceutical, supplement, cannabis or drug company. As a new venture, we are looking for the right partners who share our tireless and accountable work ethic, while engaging in and promoting a culture that cares about its stakeholders. It is time to change the way natural products are developed and to expedite real solutions that millions of people need,” he notes.
Skolnick chimes in that there is significant latent demand for non-drug solutions. “However, the dietary supplement industry’s pricing model doesn’t support the type of due diligence used by pharma. MedicascyAI speeds up certainty with less overall cost by using a set of sophisticated algorithms that analyze compounds for safety as well as efficacy in relation to the biological processes that jeopardize good health,” he concludes.
AI is increasingly being used within the nutrition industry. Last month, Danone North America entered a new partnership with Brightseed to map novel plant nutrients to human health. Meanwhile, researchers recently developed whole-body computational models to further propel research into personalized medicine, including the role of diet on the microbiome. At a consumer level, ZOE is launching a test kit and app that uses AI to develop personalized eating plans based on a person’s unique gut microbes and dietary inflammation.
By Katherine Durrell
