AsianScientist (Jun. 6, 2016) – Data scientist Li Yingrui wants to capture the massive troves of healthcare data available to us, and with it, ‘digitalize’ life. Li believes that we can someday build comprehensive, computerized versions of ourselves to predict diseases down the line—and in doing so, prevent them.
Li, who previously played an instrumental role in establishing the Beijing Genomics Institute (BGI) as a world-class bioinformatics research and services center as the company’s chief scientist, co-founded iCarbonX with BGI ex-CEO Wang Jun and Boston Consulting Group ex-partner Chun Wu just last year.
iCarbonX aims to build a digitalized data ecosystem for each and every person, based on a combination of the individual’s biological, behavioral and psychological data, and make sense of it all using artificial intelligence and data analytics. This slicing and dicing of data allows healthcare companies who use iCarbonX’s platform to provide their patients with targeted, personalized treatments.
Already, the company is making waves: iCarbonX recently completed a Series A financing round of over US$100 million, which gave it a post-money valuation of approximately US$1 billion. Chinese tech giant Tencent is one of their major investors.
Asian Scientist Magazine caught up with Li, who was in Singapore for the Forbes Under 30 Summit Asia, where he was being honored on the magazine’s inaugural 30 Under 30 Asia list.
- What is the mission of iCarbonX as a company?
In our industry, we just want to help people to live longer and healthier. It’s always painful to see people who are suffering in their last few days. The question is, how can we change this and do something that helps them substantially, and in a scientific way? The goal then is set: can people live for very long and die swiftly, rather than die gradually?
Life is complex and complicated. Descriptions of a human being—we may be talking about genes, proteins, biological regulatory networks, hormones—are already being recorded by mobile devices as well as medical records.
We now have the opportunity to document or record people in a way that we call ‘digitalized life.’ After we make a digitalized form of you, can we use that to describe your healthcare condition and status? That is why we started iCarbonX.
- What sort of ‘omics will you be testing?
We are just following the central dogma of biology, or life. You have genes, DNA modifications, epigenetics, transcriptomics, coding or non-coding RNA, and so-called metabolomics, where one analyzes small molecules.
We are collecting data from medical records, medical imaging and mobile healthcare devices. We are also collecting behavior data using different applications on your mobile phone.
- How can iCarbonX help a person with a chronic disease or, say, late-stage breast cancer?
The goal for us is to prevent it. It’s a process—we would like to manage all those risk factors from the very beginning, track the trends to see whether you are going to have a high risk of developing certain diseases, and produce a description of your status. We can then try to make some interventions, to shift the trend to get you a little bit further from the risk of developing cancer.
Healthcare and medicare are different. Medicare is when you are already broken, when you have cardiovascular disease, or when your blood vessels are damaged or calcified. But that’s not something that happens overnight. You have to be in the wrong status for quite a long time before you realize it.
The key is, can we understand what status you are in well before it develops into symptoms? That is something we want to use digitalized life to do. We want to model human beings; we want to use artificial intelligence and data engines to predict trends, and most importantly, give you personalized treatment.
- What kinds of artificial intelligence technology are you using at iCarbonX?
In terms of developing the data platforms, it’s mostly done internally, in house. It’s very important to make sure that we synchronize all the efforts because we cover so many areas.
But for the applications side, and also the data generation side, we keep it open. We hope that different companies will dock their applications or their data directly with this platform and share the data. It is not open source as it is still proprietary, but the platform and the model itself is open.
The key thing is we’re not doing diagnostics, but we enable diagnostics companies to apply this comprehensive model for clinical decision support, for example.
- How do you ensure privacy for patients who supply you with the data?
There are solutions for how you separate data management from data operating or usage. For example, people who have access to the data files may not understand them. Even if they copy the files, it’s all encrypted. It’s also not stored per individual; it is ‘mixed up.’
Secondly, it’s very important to respect regulation standards and consent. We are not asking people to give out their data; we are asking them to license their data to an application they would like to use. So it’s very different. The data is owned by you and you are licensing it to that application you would like to enjoy, which will access the data via the application interface. With informed consent, people’s data will only be used when they know it.
- What would you say is the greatest achievement of your career so far?
If you ask which achievement I consider the biggest in the last ten years, I have to say it’s not about papers or publications. It’s the one day when I was at a public welfare event as a volunteer, where I saw people in their last few hours or days. It really hit me hard, because before that I was pretty proud of myself, that there’s not many people who could carry out research at my age.
After that, I began thinking in a very different way. Can we build infrastructure or a platform that does not just enable scientists and experts, but also the general public and people who serve the general public, to articulate a better way of healthcare management? Can we enable them to enjoy the things that we promised in all those publications we published?
- What are your plans for iCarbonX over the next 10-20 years?
In my mind, the company will have several stages. The first stage is gathering and integrating all types of data, and developing relationships with businesses, applications and data generators. The second stage, I believe, is to build models for human beings based on the data. I think that’s something that will happen in a ten-year time frame.
This article is from a monthly series called Asia’s Rising Scientists. Click here to read other articles in the series.
Copyright: Asian Scientist Magazine; Photo: Nurfilzah Rohaidi.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.