BGI Hong Kong has big plans and hopes to sequence all the genetics in the world.
By Allison Proffitt
September 28, 2010 | HONG KONG—I am collected at the Hong Kong airport by Employee Number 15. Emily Yum got her master’s in physics and started her first job about six months ago with BGI Hong Kong, the recently opened office of the Beijing Genomics Institute that will handle most of BGI’s international genomics business.
Hong Kong is blanketed by a thick haze in early September and the sky matches the concrete buildings in the Tai Po Industrial Estate in Hong Kong’s New Territories. The estate is older than the Science and Technology Park, but here BGI is free to own the building. I don’t realize we’ve arrived at BGI until Emily points out the small sign, hanging low by the front door. A blue tarp on the top corner of the building suggests that signage more prominent and appropriate—given that this will soon be the largest genome sequencing center in the world—is on the way.
BGI Hong Kong’s executive director Alex Wong couldn’t be more friendly. He welcomes me into his office on the mezzanine level overlooking a completely empty loading dock.
The building was a former printing press and BGI is only the third owner, paying cash in July although extensive renovations began in late December 2009. The building got a new air conditioning system, an extensive video surveillance system, and a new power generator. BGI plans to add another power generator to meet the heavy expected power demands.
Last January, BGI bought 128 Illumina HiSeq 2000 machines, but will have 135 by the end of 2010. About 20 of those will be at BGI’s Shenzhen headquarters; the other 115 or so will be installed in Hong Kong. Four new sequencers arrive weekly. There are 46 in house the day I visit—four are still wrapped in plastic. Things are growing so quickly at BGI it is hard to keep up.
BGI will keep its sequencing fleet busy with what Wong calls a “mosaic” of commercial service contracts, R&D, collaborations, and partnerships. He expects about 80% of BGI’s work to be commercial fee-for-service contracts, the rest to be BGI-led R&D. Within the service contracts though, there will be a sliding scale. “People ask me what the unit price for a run is,” Wong says, shaking his head. It’s just not that simple.
The price will depend on the study’s complexity and BGI’s relationship with the partner. Academic and more collaborative partnerships will cost less than commercial services. BGI’s mission is to “help find out the human secret,” and the institute is open to collaborations with anyone on “all the genetics in the world,” Wong says. “If the project is an open protocol, we’ll be generous; but if the client is using the results to make money, of course we should make the service profitable!”
BGI has already caught big pharma’s attention. Wong mentions discussions with GlaxoSmithKline, Pfizer, Lilly, and Novartis, but doesn’t hint at the 5-year Merck deal announced two weeks after my visit (see, p. 42).
BGI is considering a species genome project, monogenetic disease genome projects, and clinical genetics collaboration with several academies and health authorities in Hong Kong, China, and other countries.
BGI is also expanding its own database of genetic information and will grant licenses to anyone doing human research—whether the license is free or not depends on how altruistic the research. “If the project is good for the world, we want to contribute,” he says, and acknowledges that BGI has a long way to go to achieve full market penetration. “But we will do our best to become a major sequencing partner in the worldwide market,” Wong says.
And not just partner, but provider. “Personalized sequencing will be one of our targets—absolutely,” Wong says. It falls under BGI’s mission of promoting general health. In addition to offering lump sum sequencing to hospitals and health departments, BGI aims to market the service to individuals as well. By the end of the year, he expects he will be able to do an individual’s genome for about $20,000, but the goal is a personal genome for around $2,000.
BGI’s goal is to be “the best service provider, the best partner” for genomics research, says Wong. That means delivering the highest data quality and a very strong informatics team. For instance, Emily works on BGI’s troubleshooting team, which reviews data pouring off the Illumina HiSeqs for any sign that BGI’s high quality standards are not being met.
Bar codes track which sample was processed on which instrument. At every phase, a team member checks the protocol while a colleague follows it. Cameras dot lab ceilings, each focused on one instrument, tracking which team members interact with the sample, when, and exactly what is done. Any deviations in temperature and humidity sound an alarm and “people come running”. If a customer has a question about his or her data, BGI can provide both video and written records of every variable concerning the experiment.
Giving customers the best data quality means keeping up with advancing technologies. “BGI is not only using Illumina, but they are our best partner,” Wong says. BGI does have some SOLiD instruments and has had discussions with Pacific Biosciences for a beta testing unit. The question is whether PacBio can make the system “robust, reliable, and high capacity enough” for BGI.
Wong expects a new generation of sequencers every three to five years and knows that large scale is the best way to save cost. I ask if Wong already has a favorite among the future sequencing technologies, but that is apparently a secret.
Wong says the Illumina machines are currently producing 200 gigabases per run, and expects a higher throughput by the end of 2010. At 40 gigabases per day per machine, he expects to be generating 3-4 terabases daily by year’s end.
The BGI Hong Kong supercomputer currently has 2,400 cores, but Wong and a colleague seem uncertain about the storage capacity. After some back and forth in Cantonese, they settle on a current total of almost three petabytes (PB) of on site storage in Hong Kong (Shenzhen has a little over 7 PB). The calculation capacity is about 25 teraflops. With the scope of work BGI is hoping for, that may not suffice for long.
BGI doesn’t plan to do a lot of long term storage for its clients. Some biotechs will want all of their data back, and will ask that BGI not keep any records. Others will prefer that BGI retain the data for a few years, depending on the client’s intellectual property policies. BGI will have its own policy, but will work with clients to meet their needs.
Although DNA extraction from blood and tissue samples is currently handled in Shenzhen, by the end of 2010 all international projects will be done in Hong Kong (though some processing may continue in Shenzhen). Wong expects around ten instruments to be housed abroad (BGI Europe in Copenhagen and a yet-to-be-announced office in South Asia or Australia) for small projects.
Today there are 77 employees in Hong Kong, growing to 100 by the end of the year and eventually 200 laboratory and 50 administrative staff. When renovations are completed, BGI Hong Kong will occupy more than 10,000 square meters, more than half of which will be lab space with 500 square meters of IT space.
Currently, most of BGI’s bioinformaticians will remain in Shenzhen, though some will come to Hong Kong to serve the international market. Part of Wong’s personal directive is to focus on education and develop bioinformaticians. BGI plans to host a Master’s or post-graduate course and “train up experts,” Wong says.
Asked about the challenges to BGI’s grand vision, Wong smiles. “We spend a lot of money! It’s a challenge.” Financially BGI is very healthy, he says seriously. “We have good business and we have support from some banks.”
The other challenge is conceptual. “People don’t understand preventative medicine. When a person feels well, they don’t want to do anything,” he says.
BGI is privately held and employs 3,000 people now across five centers in mainland China (Shenzhen, Beijing, Hangzhou, Shanghai, and Guangzhou), and the three existing international centers. In addition to sequencing and bioinformatics, other areas of focus include diagnostics, biofuels, and agriculture.
“Is there anything you won’t do?” I ask. Wong says BGI will try to avoid any bioethical minefields. “We’re not that crazy,” he says, almost acknowledging that BGI’s long list of goals and custom approach sounds quite daunting.
Privately, Wong promised his wife two vacations a year. So far he hasn’t delivered. “She wants to know where her trips are!” he says. But there is too much to do. “Maybe after the second phase of renovations are done next year I can squeeze out a week or two.”
This article also appeared in the September-October 2010 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.