Eagle Genomics says a combination of AI and graph is helping its clients respond to growing scientific interest in the role of microbes in maintaining health and wellbeing.
In October last year, Bill Gates delivered a speech in Cambridge to mark his receipt of the 2019 2019 Professor (Stephen) Hawking Fellowship. In it, the Microsoft co-founder developed a set of wide-ranging predictions about the future of healthcare—including the potential of what we could do if we better understood “the community of good bacteria that live in your body,” the microbiome: “Until recently, fixing the microbiome has been a complete mystery to us [but] over the next 10 to 20 years, we’re going to learn more about each individual microbial species and how they work with the food you eat to impact health. [And] that knowledge will allow us to smartly engineer interventions that ‘correct’ the microbiome when it’s out of whack.”
Healthier
If he’s right, that means, potentially at least, we might have techniques to not just make our food healthier, but new ways to combat non-communicable diseases like non-alcoholic fatty liver disease, obesity—maybe even some forms of cancer. Nodding along enthusiastically in the audience: Anthony Finbow—the CEO of a UK biotech company that is trying to use AI to do just that.
Specifically, he and his team over at a from called Eagle Genomics are building tools for researchers at food and wellness companies to help them explore vast, complex biome data using the Microsoft Azure Cognitive Services stack, as well as network data approaches like graph. “We’re trying to apply data science principles and structures to industrialise approaches to life science data exploration,” Finbow told RockingRobots. Essentially, that means enabling automated and cataloguing of biome data and metadata and then presenting a user interface to enable the surfacing of connections between entities of interest—e.g. the nodes and and edges of the graph—to be exposed so scientists can get a visualisation of the nanoscale organisms that they’re studying.
What you get back, he claims, is a biome-focused version of what Gartner’s called the ‘data fabric’ for enterprises to unify and navigate this data based on context—allowing them to ask questions and explore interactions between microbes in the host off an AI-powered landscape where scientists can range and navigate to the most appropriate data for their analysis.
Maybe this toothpaste really is good for me
Sounds pretty abstract? Finbow offers a use case—working out scientifically what causes some of us to put on too much weight. “We don’t understand which food ingredients drive obesity; we don’t know how to transform foods to reduce the epidemic. Food companies are trying to understand that, so they need to understand the active ingredients and the properties of their foods. But they also need to understand which microbes in your gut are transforming these ingredients—what they are producing, and how that the substances that produce are of benefit or disadvantage to the host in exacerbating these inflammatory disease challenges, or in ameliorating the challenges.”
Challenges
Investors are beginning to think both Gates and Finbow are on to something. Late last month the company closed $9 million in new scale-up funding to further develop the capabilities of this so-called ‘knowledge discovery platform,’ e[datascientist], with money coming from a consortium including network researcher Albert-László Barabási. Brands are also getting convinced, too: Unilever has already used the platform successfully to distil credible scientific evidence to support claims about its Zendium toothpaste, which uses natural enzymes to boost the good bacteria in the mouth—the first substantiated microbiome-based claim put out as part of a consumer wellness product launch.
“If we can enable those big food companies to solve those challenges, we’re on a truly exciting path,” he concludes. Let’s see if he and Bill have spotted a real winner here.
