Ex-BGI CEO sets up AI-enabled health data mining operation

Jun Wang

The former CEO of Chinese sequencing giant BGI has started his new venture, iCarbonX. As expected, iCarbonX will build a deep, broad pool of health data and develop an artificial intelligence platform to dig into the resource for insights into how to treat diseases.

Details of iCarbonX gleaned by China Daily at the official unveiling of the business in Shenzhen, China, suggest ex-BGI CEO Jun Wang is following the plan he shared with Nature News shortly after leaving his previous post. The vision presented at the launch party was of a business built upon a repository of proteomic, metabolomic, microbiomic and other omic data, which will be paired to phenotypic, social and environmental information. Wang thinks that interrogating such a resource will lead to advances in how individuals manage their health and biopharma companies develop drugs.

"We want to establish a health-related big data platform, to develop artificial intelligence to interpret and mine the data as well as to enable every individual to better manage their health and defeat diseases," Wang said. To deliver a return on the outlay needed to set up the operation--Wang said in July that the initiative may require $1.6 billion--iCarbonX will open up its data to drugmakers, hospitals and healthcare service providers. From its base in Shenzhen, the home of BGI, iCarbonX intends to serve clients around the world.

Wang has enlisted the support of some notable names for the project. Yingrui Li, the former chief scientist at BGI, is a co-founder, as are Chun Wu and Hao Li, who are best known for their stints at Boston Consulting Group and China Unicom, respectively. Wang is tapping Hong Kong University of Science and Technology's Qiang Yang for artificial intelligence know-how, while Karolinska Institutet's Qiang Pan-Hammarström and Cornell University's Zhenglong Gu are providing expertise in cancer genetics and nutrition.  

China Daily compared iCarbonX to a Google ($GOOG) for biotech. But in many regards the model is reminiscent of work already underway at Lee Hood's Arivale, J. Craig Venter's Human Longevity, and Google's own Calico. Each is placing a bet on the idea that gathering lots of data, from an array of sources, and digging into it will lead to breakthroughs in human health. 

- read China Daily's article