Nearly two years ago, the San Diego-based genomics giant Illumina reaffirmed its dominance of the DNA sequencing market when it announced a fast and powerful new machine, called the NovaSeq, that’s since been adopted in labs all over the world.
On Thursday, at a genomics conference in China, a Chinese company called BGI unveiled its own new machine. That machine, the company said, can sequence a whole genome in 24 hours. It can generate six terabytes of sequencing data — about as much data as a person would use watching Netflix nonstop for half a year — over that period. And, again, according to the company, it can do it all with the utmost precision.
All of which could, at the right price, make BGI a very fierce competitor for Illumina — and possibly drive the already-plummeting cost of sequencing down even further.
“It seems that the race is on, and geneticists everywhere are the big winners,” said Christopher Mason, a geneticist at Weill Cornell Medicine.
The race is being closely watched by scientists and clinicians. Cheaper, faster sequencers, after all, make it possible to generate more medical knowledge.
But BGI’s promises have traditionally been viewed with skepticism in the U.S. And even if Illumina is the Goliath of the U.S. sequencing market, BGI makes for a unusual David figure, not least because of a crackdown on overseas investments and a brewing trade war between China and the U.S.
There is little doubt that BGI is a major player in the global market. The company, headquartered and publicly traded in China, sells its sequencing machines in Asia, Europe, and Canada. And it counts the American genomics pioneers George Church and Leroy Hood among its scientific advisers.
The company also runs a genomic services business. Its customers — which include academic researchers, drug companies, and third-party sequencing providers — pay to ship DNA samples to BGI, which then sequences them and sends back the genetic data. BGI pitches its services to customers both in the U.S. and overseas. But it’s not clear how large BGI’s services business is in the U.S.
The company says that MGISEC-T7, its new machine, is not just an upgrade to any of its existing sequencers; it’s a new product meant to serve customers looking for highly accurate and thorough analysis of DNA.
The sequencing capacity is comparable to the capacity that Illumina originally announced with its NovaSeq line of machines, though Illumina has since presented data suggesting that its machine can approach 10 terabytes of sequencing data daily.
Mason called BGI’s announcement “intriguing,” though he cautioned that the company will have to get its sequencers physically positioned in the U.S. market before scientists begin using them en masse.
(BGI covered the cost of sequencing the DNA samples in an analysis of BGI’s data quality that Mason ran recently, and Illumina has paid for Mason’s travel in the past.)
BGI has not yet disclosed where it plans to sell its new machines or how much it will cost to use it to sequence a genome.
When it comes to the cost of sequencing, “making it affordable to everyone around the world is our core mission. That’s not the core mission of our competitors,” said Kristi Heim, a spokesperson for BGI.
Sequencing a human genome with Illumina’s most cutting-edge machines is estimated to cost around $1,000. But it’s hard to directly compare the costs of sequencing with Illumina’s and BGI’s machines, for a few reasons.
For one, Illumina has a different type of business than BGI. While Illumina does some servicing, it makes money primarily by selling things: mostly reagents and other products used in running the machines — as well as the machines themselves, at price tags nearing $1 million a box — to sequencing providers like the Broad Institute in Cambridge, Mass.
Those sequencing providers often give substantial discounts to customers who order in bulk; not unlike shopping at Costco, you can get 1,000 genomes sequenced at a cheaper per unit rate than you can a single genome.
What’s more, there are lots of factors that go into the ultimate cost of sequencing a genome. Among them: the upfront costs of buying a machine. The chemicals it takes to run the machine. The overhead expense of paying technicians to run it. The salaries of the scientists who can interpret the results. Electricity bills factor in, too.
An uphill battle
As geneticists watch BGI for signs of whether it may soon mount a major push into the U.S. market, it has already begun to make moves.
It has offices in San Jose, Calif., Seattle, and Cambridge, Mass. And it has academic partnerships with some of the country’s leading medical institutions, including the Mayo Clinic and the Children’s Hospital of Philadelphia.
The company earlier this year reached an agreement with Toronto’s Mount Sinai Hospital to install BGI sequencers there — the company’s first in North America — as part of an effort to develop a way to test pregnant women for their risk of pregnancy complications and premature birth. (The company already sells prenatal tests to Chinese parents-to-be.)
Still, BGI faces big obstacles that could continue to keep it from competing in the U.S. Illumina has strong patent protection and a reputation for defending its intellectual property.
And, like other Chinese companies, BGI must persuade U.S. customers that they can trust the quality of the data it generates — and that it’s a trustworthy custodian of sensitive health data. In the past, the company has been viewed with suspicion and called an arm of the Chinese government. (The company was built up with state support and in the past few years has continued to receive modest government subsidies, equivalent to seven figures in U.S. dollars.)
In China, “trust about quality and intellectual property is a moving target. So BGI has struggled with its reputation, I think, as an entity that has innovated through its own technological R&D,” said Mark Massaro, an analyst who covers Illumina for Canaccord Genuity.
Massaro pointed to BGI sequencer-generated data sets that the company recently posted online, which one independent analysis found to be comparable, if slightly lower in quality, to data generated by Illumina’s machines. “Can we trust that this data is real? I don’t have a lot of faith in it,” Massaro said.
“I think those of us who have been in the space for a while, we tend to wait to see the data in our hands before we get too excited about anything,” said Sean George, CEO of Invitae, a medical genetic testing company headquartered in San Francisco. (The company mostly uses Illumina products, with a few exceptions.) “There’s always promises about the next new platform.”
Illumina itself has stoked concerns about the quality of data from BGI’s proprietary machines.
“If you compare what customers can get with our products compared with what they can get in BGI, there’s a big difference in terms of the quality of the data that you get off the instruments, in terms of the reliability of the instruments if you purchase the instrument yourself,” Illumina CEO Francis deSouza said Tuesday during the company’s quarterly earnings call.
What’s more, among companies that are buying their own sequencers, Illumina may be impossible to dislodge from its foothold in the marketplace. The cost of sequencing has been declining each year, leading some to compare Illumina’s position to Intel’s early dominance in computing.
“Every year, Intel brought out faster chips that did more processing, cheaper. And Illumina is doing very much the exact same thing. Every year, more sequencing capacity comes online at lower prices,” George said.
Intel’s command of the market created a consistent architecture — and Illumina may be doing much the same thing, he noted. If companies that have already invested in Illumina’s machines can lower the cost per sequence without overhauling their workflow, “that’s a pretty powerful combination.”
A history of ‘starts and stops’
Like the broader Chinese economy, BGI got its start as a government-funded operation before growing into a global powerhouse.
Founded in 1999, it was originally called Beijing Genomics Institute, because it was run out of Peking University. BGI was set up to play a role in the Human Genome Project; it ultimately worked with other Chinese scientists to map out 1 percent of humanity’s genetic blueprint as a contribution to the broader international effort.
After that project, BGI turned to building a real business. It started investing heavily in research, sequencing a variety of different genomes, including the rice plant, a panda bear, and a 4,000-year-old man.
By 2007, BGI had moved to its current home of Shenzhen, just across the border from Hong Kong. In 2010, BGI signaled its ambitions when it purchased 128 Illumina sequencing machines (then a state-of-the-art model called the HiSeq 2000).
Not long after, BGI tried to acquire Complete Genomics, a Bay Area sequencing company that was at the time something of an emerging rival to Illumina. Some in the U.S. were wary of the planned deal, suspecting that the Chinese government wanted access to American technology. Illumina, which had failed in its own bid to buy Complete Genomics, even hired a Washington lobbyist to try to generate opposition to the deal. But BGI’s deal went through anyway, in 2013.
In the half decade since, BGI has worked to improve Complete Genomics’s technology. It launched a line of sequencers, known as Revolocity, that quickly failed. The more recent lines of BGISEQ sequencers seem to be doing better, but traction still has been limited.
“BGI has had a lot of starts and stops — more stops than starts,” Massaro said.
The company has managed to push past those previous stops — but political headwinds may make its path more challenging in the future.
BGI’s own annual report indicates that uncertainty around the relationship between the United States and China may pose a risk to the company.
And Chinese sequencing companies are clearly worried about the potential impact of tariffs on their ability to get some of the raw materials they need from American companies like Illumina. Over the last two quarters, a handful of Chinese companies have bought more than $27 million worth of extra sequencing supplies, Illumina reported Tuesday.
Despite the challenges facing BGI, the company has real promise, said Serge Saxonov, CEO of 10X Genomics, a Bay Area company that sells technology meant to give labs a richer picture when they sequence DNA. When it comes to building sequencers, Saxonov said, “execution really matters over the long term.” And for BGI right now, he said, “it’s relatively early days.”
Correction: An earlier version of this story incorrectly described the largest driver of Illumina’s revenue. The company makes most of its money selling reagents and other products used in running its sequencing machines; the machines themselves generate less money.
“Founded in 1999, it was originally called Beijing Genomics Institute, because it was run out of Peking University.”
This statement was not true, check it even for your opinion piece. I was amsued so many time, West media just cannot get the whole true stories.
The statement that BGI originally was run out of Peking University is not accurate. It never associated with that university, but Chinese Academy of Sciences, in the suburb of Beijing.
scientific advisers and scientific advisory boards, are just people paid for using their name and photos on company websites. I.e., worth less that a professional athlete actively promoting sport shoes.
But apparently, effective enough to fool investors and media in granting trust.
Comments are closed.