On 5 January, Shenzhen-based iCarbonX held a one-day summit to announce its much-anticipated plans for an artificial intelligence (AI)- based health care platform. The big news was an alliance uniting iCarbonX with seven other companies, stitched together with over $400 millions of investment and with iCarbonX’s visionary founder bioinformatician Jun Wang as the linchpin.
Wang is known for his blazing record of publication when he was the CEO of Shenzhen- based BGI Genomics, a post he left last July. Of his new endeavor, Wang says it will be the fastest-moving of all the big data personal health companies.
The alliance’s goal is, first, to pool results from genomics, proteomics, metabolomics, transcriptomics and immune response, along with analyses of gut bacteria and lifestyle factors; and then to analyze the resulting unwieldy data sets with algorithms produced by iCarbonX. Users will access their results through a mobile phone application. At the summit, Wang demonstrated how the application, Meum—meaning ‘my’ in Latin—dispenses advice on nutritional, exercise and sleep needs. Eventually, the aim is to mine an individual’s data to provide information on the person’s susceptibility to disease linked to actionable advice.
Five US companies form the alliance along with Robustnique, a cosmetics company based in Tianjin, and Tel Aviv–based Imagu Vision Technologies, acquired by iCarbonX in September. US-based partners include SomaLogic, General Automation Lab Technologies, HealthTell, AOBiome and PatientsLikeMe. Overall the alliance members account for 1,000 staff, 200 of them are at iCarbonX.
“We’ve thought about AI forever, but Jun crystallized it,” says Larry Gold, founder and chairman of alliance member SomaLogic, of Boulder, Colorado. “People here have a real interest in making democratically priced omics data a reality as soon as possible,” said Gold during the summit.
Many big players, including IBM and Google, are already pursuing similar AI-driven approaches to healthcare, but iCarbonX, by bringing in expertise from all alliance members, expects to develop faster and cheaper tools for data production, says Wang.
SomaLogic, for instance, will contribute the proteomic platform SOMAscan, which, the company claims, can quantify some 4,200 proteins in a sample; San Ramon, California– based HealthTell uses silicon-wafer-based microarrays that hold 330,000 peptides to rapidly detect antibodies and give insight into a person’s disease state as well as drug or vaccine responses. Wang says monitoring of individuals with the iCarbonX platform will be comprehensive, including frequent proteomic and antibody sampling and 24-hour glucose monitoring, and affordably priced. “Some companies are doing one or a few of these things. No one is assembling it all,” says Wang. The group is also training coaches to help users digest the information.
iCarbonX has so far accrued over $600 million in investment. The original $200 million came within a year of its October 2015 founding from Shenzhen-based Tencent, which produces WeChat, China’s social media application with some 800 million users; Zhongyuan Union Stem Cell Bioengineering Company, a biotechnology company based in Tianjin; and the Tianfu Group, a Shanghai-based consulting group. iCarbonX is not divulging the source of the new $400 million that it plowed into the seven alliance members. $7.5 million went to HealthTell, but other amounts are undisclosed.
For each partner, the agreement and investment are different. For SomaLogic, the company will provide proteomics expertise to the alliance; iCarbonX will invest (an undisclosed amount) in the company and help to set up a SomaLogic joint venture in China. But some details need to be worked out. “We’re still dancing,” says SomaLogic founder Gold, who hopes it remains a tight alliance. “[Wang] is someone you’ll follow gladly. You don’t meet many people like that. I’m all in.”
PatientsLikeMe co-founder and chairman Jamie Heywood, left, with iCarbonX founder Jun Wang are partnering to accelerate a deeper understanding of human health and disease.
Jamie Heywood, co-founder and chairman of Cambridge, Massachusetts–based PatientsLikeMe, says he is “energized” by the alliance and by Wang’s laissez faire approach. “Jun didn’t put on any constraints. He is giving us an opportunity to go further and faster than we could anywhere else,” says Heywood, whose company is an online portal where some 500,000 registered patients submit clinically useful data about their experience with drugs and disease. Wang plans to use PatientsLikeMe as a base for getting into the US and to create a PatientsLikeMe type portal in China. “The beauty of this ecosystem is that it’s good for all of us,” says Heywood.
The alliance will have to convince some observers of its uniqueness. Michael Snyder, director of the Stanford Center for Genomics and Personalized Medicine, sees the Chinese- led project as just one of many—including Seattle-based Arivale, San Diego based Human Longevity Inc. and his own initiative, Integrated Personal Omics Profiling— that are adding other omics to genomics in an attempt to tailor medical knowledge to individuals (Nat. Biotechnol. 3, 887–889, 2015). “We already know that the genome is just a small part of the equation and that other types of data can be quite valuable,” says Snyder.
Snyder admits, though, that the investment is impressive. “Its significance will depend on how many of people they follow and the quality of data they can garner,” he says.
Wang plans to gather one million users within five years, a number that will give the alliance an edge, he says. The plan is to tap health check-up clinics as well as insurance companies in the region. Representatives from the two largest Chinese insurance companies, AIA Group and the China Life Insurance Company, were at the summit. China’s size and culture—which already includes an emphasis on preventative care through traditional Chinese medicine—makes Wang confident of hitting his target. Critically, the alliance plans to recruit through the wildly popular social media application WeChat. “I will try everything and go with whatever works. No one will be able to collect at the same scale of data that I am doing,” he says. “A million is a starting point, not an endpoint.”
The numbers are crucial if one is hoping to get through the significant noise in such a weighty data set to find strong signals, says Todd Krueger, president of AOBiome, another alliance member based in Cambridge, Massachusetts. “It’s the value density. The informational value from 50 patients to 5,000 patients to 500,000 patients rises nonlinearly.”
Still, the alliance will probably start modestly, says Mark Natkin, managing director at Beijing-based Marbridge Consulting. Like Arivale, a systems biology wellness company founded by Lee Hood that has just launched a personalized weight loss program, iCarbonX may begin with nonmedical products. “It may succeed, but will also likely hedge its bets, leveraging the platform to offer personalized cosmetics or healthcare products that don’t really have to stand up to clinical tests for efficacy.” Precision skincare is one of four fields of focus for iCarbonX, along with precision nutrition, precision health and precision medicine.
Because many alliance members are already experienced in mapping omics data to disease and treatment choice, precision medicine will make an easy starting point. Rhoda Au, a neurologist at Boston University and member of the Framingham Heart Study who is discussing potential collaborations with iCarbonX, says the alliance also seems ready to take on the newer and more challenging field of precision health, by bringing health monitoring to people’s homes—including smart toilets (for analyzing urine or feces), smart mirrors (to map changes in body shape to aid health diagnosis) and smart robots (a chat tool). Upstart iCarbonX has an edge in this nontraditional field of precision health, Au notes, because as a new company it is willing to tap its investors for resources to drive such projects. “What we are proposing to do is not something I could raise using traditional research funding mechanisms (e.g., foundations and NIH).” She adds that, on the user interface side, in addition to the app, “I like what they are doing with their smart robots—that’s a good delivery system for those less screen inclined.”
Being in China is also a major perk. China is investing heavily in research and healthcare and incentivizing opportunities to increase domestic spending so as to reduce its reliance on exports. “iCarbonX can tap into talent and resources from China in a way that US companies cannot,” says Au. “And in the world of computational science and technology, this is a big advantage. I see iCX as being able to maximally capitalize on their unique positioning.”
Snyder does suspect that laxer regulations on interactions with patients in China will
give Wang’s alliance
room to breathe that
US competitors lack.
“This space is highly
regulated in the US. It is difficult to get permission even in a research context to return participants’ information back to them,” he says.
Natkin agrees that having its center of gravity in China will be strategically important. “China is looking to foster the development of local champions,” he says. What’s more, over the past few years, China has tightened restrictions on the export of human genetic resources, and as a result, foreign rivals attempting to gain traction in China “will face a variety of obstacles making it difficult to compete effectively in the China market,” says Natkin. In China, there is still a large ‘gray area’ between what is considered a diagnostic, which requires drug agency approval, and what is just health advice. “We will find out,” says Wang.
But the alliance, which will also be operating in the US, will still have to deal with regulations there. Wang says, however, that PatientsLikeMe has learned from the US Food and Drug Administration’s clampdown on consumer genetics testing company 23andMe, of Mountain View, California, for diagnosing patients without having validated the tests, and has learned to stay out of the FDA’s crosshairs.
In attempting to integrate data from US patients, such as those on PatientsLikeMe, and from Chinese individuals to raise numbers, the alliance will also have to face the stricter Chinese restrictions. Can individuals post their own data online or in social media? Do the restrictions apply to proteomic data as well as genetic? “These are good questions. Regulations are a moving target and we need to understand,” Wang says.
In the meantime, Wang will be setting up a Chinese version of each of the seven companies in the alliance. He also has two pilot projects. In one, 100 people will pay RMB10 million ($1,440,000) over ten years for a package that includes life insurance, health insurance and medical insurance along with the whole range of tests and tools that the alliance has to offer. In the other, 1,000 people will be allowed to purchase one time ‘check-ups’ at pricing platforms ranging from free to RMB99,999 ($14,500)—and varying by the number of tests (immune system measurement, proteomics of different tissues, etc.) and varying degrees of information for each of those tests (for example, different number of proteins measured).
“The initiative is exciting and cutting edge, and there is no doubt that it will produce very interesting and cool data,” says Jing Ma, a cancer epidemiologist at Harvard. But the alliance seems to lack sufficient ties with the medical community to evaluate health and disease outcomes, Ma adds. The iCarbonX platform might also prove too time-consuming and expensive for people who are not sick to use.
Wang acknowledges the uncertainties—in the science, in the regulations and in peoples’ reactions toward technology—and wants to tackle them head on. “That’s the beauty of it—the challenge and opportunity,” he says.
David Cyranoski Shanghai