China's Sequencing Powerhouse Comes of Age
Damien Marsic
for those who are not afraid of the future : http://www.rael.org
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Source: http://www.sciencemag.org/content/335/6068/516.full
China's Sequencing Powerhouse Comes of Age
With new sequencing centers in Europe and the United States, BGI hopes its growing clout will help deliver the benefits promised by genomics—and revenue to pay off a mounting debt.
Such dealmaking is a central element of BGI's improbable emergence as a genomics superpower. Unlike other major sequencing centers that are attached to institutes with inhouse biology expertise and long-standing scientific programs, BGI, since its inception in 1999, has focused on developing its sequencing and bioinformatics capabilities while turning to outside teams in crop science, human disease, and microbiology to help define its research objectives.
As a result, BGI has become the go-to organization for groups in need of sequencing. At the meeting, BGI inked deals with Johns Hopkins University in Baltimore, Maryland, to work on synthetic yeast and with the International Rice Research Institute (IRRI) in Los Baños, Philippines, to study genomic diversity of 3000 rice strains. Earlier last year, BGI agreed to collaborate to sequence 5000 cassava varieties and 10,000 autism patients and family members, and do heaps of sequencing for the international Human Variome Project (Science, 16 December 2011, p. 1487). And it's looking for partners on its own audacious goal to sequence a million human genomes, a million plant and animal genomes, and a million microbial genomes. To foster closer collaboration, BGI is establishing two sequencing centers in the United States and one in Europe, with more in the offing. The growing web of partnerships is turning BGI into the only genomics enterprise with a global footprint.
In building its empire, BGI has made a huge gamble. To fuel its growth, BGI has borrowed heavily from a $1.5 billion line of credit from a government-owned bank—debt it must start to pay back in 10 years. Elsewhere, governments have invested in genomics and sequencing centers with the expectation of future payoffs in medicine and other areas. But none has taken as entrepreneurial an approach as China has in counting on a single institute to recoup its investment.
Competitors are puzzled by the BGI phenomenon. “I don't think their model is very different in terms of interactions [with researchers],” says Richard Gibbs, director of the Human Genome Sequencing Center at Baylor College of Medicine in Houston, Texas, one of the big three sequencing centers funded by the U.S. National Human Genome Research Institute (NHGRI). Applying genomics to biomedicine is “pretty straightforward” in matching sequencing technology and expertise with funding and science, he says. Others worry that BGI's sheer size and mounting debt could skew priorities across the field if sequencing projects are determined by the likelihood of a payoff. Jack Gilbert, an environmental microbiologist at the University of Chicago and Argonne National Laboratory in Illinois, collaborates with and admires BGI. Nonetheless, he worries, “If one organization dominates the market and has an agenda, this could be bad for innovation in science.”BGI officials say they are not bent on world domination. “I don't think this is the only [sequencing] model that fit s the entire field and should not be the only model,” says BGI Executive Director Wang Jun. But he adds that their approach to genomics “is probably a model no one else could easily follow.”
As a research institute sponsored by the Shenzhen municipal government, BGI has set up private nonprofit institutes in Hong Kong, Europe, and the United States to manage its sequencing centers. It owns private companies that offer sequencing and clinical services. BGI also has a college in Shenzhen it hopes to turn into a university and an annual genomics conference, and it is launching a journal, GigaScience. Wang says they are simply trying to attract sufficient resources “to prove that the genome matters.”
Seizing the day
According to BGI brass, the center's rise was happenstance. “We went one step by one step,” says Yang Huanming, BGI's charismatic chair. “I don't think BGI had a long-term strategy, but it did have long-term dreams and it has been opportunistic,” adds Gane Ka-Shu Wong, a biosystems informaticist at the University of Alberta in Edmonton, Canada, who has been a BGI associate director from the start.
Yang, Wang Jian, and a handful of colleagues formed Beijing Genomics Institute in 1999 to give their homeland a foothold in sequencing. That year, backed by seed funding from the Chinese Academy of Sciences (CAS) and promises of further government support, Yang persuaded the in ternational Human Genome Organisation to allow China to sequence 1% of the human genome. It was the only developing country in the landmark project, a fact that dominated news coverage in China when sequencing was completed in 2001.As work on the human genome wound down, BGI faced a problem: It had 150 people, sequencing machines, and supercomputers but no continuing funding. To win grants, “we had to do something else,” Yang says. They took aim at the rice genome, a project BGI initiated with Chinese partners. By the time they reported the draft sequence in Science in 2002, BGI's payroll had swelled to 300, and it was even hungrier for grants. “We decided we would take on any problem related to genomics,” Yang says.
BGI joined with domestic partners and won grants for other sequencing projects deemed important for China's economy or prestige, including the chicken, silkworm, and panda genomes. And it expanded its overseas ties. BGI participated in the International HapMap Project, among others. In the process, it developed a recipe for success in its collaborations. For example, to tackle the pig genome, BGI took charge of sequencing and bioinformatics while Danish partners handled functional analysis. Every step of the way, Yang says, they sought to drive down costs, increase speed, and cover expenses.
In late 2003, riding political support for having quickly sequenced several strains of the SARS virus, BGI became a CAS institute. But it was a poor match. “There was serious conflict between CAS and BGI,” Yang says. CAS institutes have an academic structure, with senior scientists often heading small teams. BGI was building a large-scale sequencing operation staffed with an army of programmers and bioinformaticists. More important, Yang says there was little recognition of the importance of sequencing. There was “almost nothing” in funding for sequencing in China's 5-year plan that ended last year, he says.Shenzhen's government came to the rescue. In 2007, the prosperous special economic zone, eager to expand its economic base beyond manufacturing, offered 20 million yuan ($3.2 million) per year for 4 years and found BGI a home in a former shoe factory. Conflicting visions split BGI's scientists. Some chose to remain under CAS as the Beijing Institute of Genomics while Yang and his supporters decamped to Shenzhen.
In early 2010, BGI hit the jackpot when the China Development Bank extended a 10 billion yuan ($1.58 billion) line of credit. The bank finances projects in line with central government policy, which includes support for biotech, Wang Jun says.
BGI immediately bought 128 Illumina HiSeq 2000 sequencers, e ach capable of sequencing a human genome in a week. In one stroke, BGI acquired more sequencing power than any other center in the world. Winning an arms race was not the objective, Wang Jun says. “We just did a rough calculation: If we sequence the major farm animals and sequence the major medical genomes, how many instruments do we need?” They also predicted that the Illumina machines would not be trumped by a new generation of sequencers for at least 3 years.
BGI Projects
View a map of major genomics projects BGI is undertaking with international partners.
CREDIT: D. NORMILE/SCIENCE
Massive sequencing capacity is one prong of BGI's strategy. It has also strived to blaze a trail in bioinformatics by developing free software for the community. “They've been astonishingly open in the whole [bioinformatics] effort,” says Mark Edwards, a plant geneticist at Southern Cross University in Lismore, Australia.
Bioinformatics has become increasingly important as sequencing volume has grown. For example, a BGI-Danish project called LuCamp sequenced all exomes, the genome's coding regions, of 1000 patients with metabolic disorders and 1000 controls. The project “was the first of its kind,” says team member Rasmus Nielsen, a computational biologist at the University of California, Berkeley. Earlier disease-related genomics efforts involved more limited sequencing and produced less data. Before starting, “you need to know how many samples and to what depth you have to sequence to maximize the statistical power to identify a particular gene or loci that is associated with a disease,” says Li Yingrui, who headed BGI's bioinformatics for LuCamp. The project required simulations and new statistical models on scales not previously attempted. Sequencing was completed in 2010; the first report from the ongoing functional analysis appeared in Nature Genetics in October 2010.
Li's role in LuCamp highlights another facet of how BGI does things differently. When the project started in 2008, he was 21 years old and cutting master's degree courses, preferring to work on a real-world problem. “The textbooks are all behind the cutting edge,” Li says.
When BGI put a student in charge of bioinformatics and staffed LuCamp with 20-somethings, its partners were understandably anxious. “We worried about whether they could deliver, given that most of the people working on the project were not trained scientists,” Nielsen says. But the team delivered the goods.
Li eventually dropped out of grad school—“it was kind of a waste of time,” he says—and now heads a 400-strong group at BGI exploring future technologies.
A package deal
The dramatic decline in sequencing costs, from $50,000 for a human genome 5 years ago to approaching $1000 today, was expected to spur a “democratization” of sequencing. Instead it is sparking more-ambitious sequencing plans with ever-more-complex logistics and bioinformatics. That plays to BGI's strength, which is “the integration of both the sequencing and the bioinformatics,” Nielsen says. And it explains BGI's attraction to its growing stable of partners. For the rice project, after BGI sequences thousands of rice strains, IRRI intends to link biological characteristics to sequence data—in other words, connect phenotype to genotype. It aims to identify genes for desirable traits such as drought tolerance, pest resistance, and grain quality to improve commercial varieties. “BGI is the only one with the muscle” to carry out such a massive effort—sequencing 3000 strains and possibly 7000 more later—in a reasonable amount of time, says IRRI plant geneticist Hei Leung.
Hammarström, meanwhile, hopes to unravel the genetics of immune disorders. He and others have hunted for disease-related genes by looking at single-nucleotide polymorphisms in the genomic region associated with the major histocompatibility complex (MHC), a cell surface molecule that is key to the immune response. That approach has come up empty-handed, Hammarström says. Unraveling the genetics is tricky because many conditions are rare and likely involve multiple genes, hence the need for a huge number of samples. BGI will sequence the MHCs or exomes of 100,000 patients and controls in 5 years. The plan is “ambitious, yes, but within reach,” Hammarström says.The Center for Applied Genomics of Children's Hospital of Philadelphia (CHOP) is also partnering with BGI. CHOP and BGI had originally hoped to get NHGRI funding to set up a new sequencing center. When it became clear they would not win support, the two institutes dipped into their own resources to establish BGI@CHOP. Pediatrician Hakon Hakonarson, the center's director, says CHOP is already basing clinical decisions on the likely efficacy and side effects of medications by comparing patients' genotypes with samples in the hospital's biobank—the largest of its kind, with samples from most pediatric diseases along with data on medication, response, and side effects. Through whole-genome sequencing of samples, they expect to make more comprehensive and precise clinical decisions. “This is the future of medicine,” Hakonarson says.
To execute this plan, BGI will set up a sequencing center at CHOP. It will start with 10 sequencers and could later add up to 40 more. One big advantage of the U.S. operation, Hakonarson says, is that BGI will have a better shot at winning National Institutes of Health grants. He envisions half of the sequencing capacity set aside for studies using CHOP's biobank and half for BGI's non-CHOP projects.
CHOP is one of three beachheads that BGI is establishing outside China. Over the past year, it has announced plans to also set up sequencing operations in Copenhagen and at the Sacramento campus of the University of California, Davis. Centers are being mulled for southern Europe, Japan, and Australia. Baylor's Gibbs welcomes BGI's American ventures. “We're thrilled that we've got more people in the mix,” he says. He believes the U.S. centers have an advantage over BGI in expertise integrating sequencing, bioinformatics, and biology.
Day of reckoning
BGI may not be able to sustain its breakneck pace forever. Yang says that although BGI can sequence more economically than anyone, “we do not make money on it.” Thanks in part to services such as prenatal screening and genetic typing for bone marrow or organ recipients, BGI had about 1 billion yuan ($158 million) in revenue in the year through March 2011—enough to break even, Wang Jian says. They pay interest on the debt but must start paying back principal in 10 years. To generate more income, BGI is expanding its applied research portfolio. It launched an agricultural department in 2009 to use genomic techniques to improve cultivars, particularly of overlooked crops such as foxtail millet. BGI is developing genetically modified minipigs as models for breast cancer and degenerative diseases such as Alzheimer's and atherosclerosis, and it has high hopes for payoffs from synthetic biology.
With partners in diverse fields, BGI could be better positioned than its rivals to benefit if genomics begins to produce long-promised rewards. “They're being very intelligent about picking their collaborators,” says Richard Cotton, a geneticist at the University of Melbourne in Australia who works with BGI on the Human Variome Project. “In the next 10 years,” Wang Jun predicts, “we will be able to develop something useful for ordinary people.” That may allow BGI to begin paying down its massive loan.
BGI's day of reckoning is at least a decade away. For now, its bold ambitions are carrying the day. “They make you realize the sky is the limit,” Hammarström says.