Complete Genomics (CGI)
Since it started sequencing genomes in 2009, the DNA services company Complete Genomics has garnered praise for its technology, but little commercial success. Researchers valued its ability to produce readouts of entire human genomes with minimal errors, but they feared for the future of the company, based in Mountain View, California, which lost US$72.3 million in 2011 alone. Now BGI, whose sequencing facility in Shenzhen, China, is the largest in the world, has tossed Complete Genomics a lifeline.
On 17 September, BGI announced that it had agreed to merge with the struggling firm and would spend $117.6 million to buy up its shares. The rescue was welcomed not just by the company’s board of directors, but also by those who worry that diminishing choices is a problem for genomics.
“Right now, Illumina would have a monopoly if Complete Genomics went under,” says Michael Snyder, director of the Stanford Center for Genomics and Personalized Medicine in Palo Alto, California. Instruments made by Illumina of San Diego, California, account for most of the world’s sequencing output. Maintaining competition in the industry will keep prices low and technologies improving, Snyder says.
Like Complete Genomics, BGI sells sequencing services rather than equipment or reagents, but it relies chiefly on instruments made by Illumina. Complete Genomics has its own instrument and software platform, which is specialized for sequencing whole human genomes.
Although comparing ever-changing platforms is never straightforward, a 2011 analysis of the same genome read by Complete Genomics and Illumina technology showed that although the former found fewer variants in the DNA sequence relative to a standard reference, it was more accurate (see ‘Careful reading’)1. In July, Complete Genomics described a technique that reduces error rates in sequenced genomes to 1 in 10 million base pairs and shrinks the amount of DNA required to generate a sequence2. The technique also distinguishes between DNA fragments that come from paternal and maternal versions of the same chromosome. This can reveal, for example, whether mutations occur in one or both copies of a gene, and so whether an individual is at risk of developing a certain disease or is merely a carrier.
|Single nucleotide variants||Insertions and deletions|
|Complete Genomics only||99,578||94%||206,461||100%|
*A small, random sample of variants was reanalysed by a more accurate but more expensive method.
Source: Ref. 1
“That’s going to be a game changer in how whole-genome sequences are interpreted and analysed,” predicts Thomas Barber, a geneticist at Eli Lilly in Indianapolis, Indiana, who works with both BGI and Complete Genomics. BGI’s acquisition means that researchers will continue to benefit from such advances. Having the two companies under the same ownership could combine the strengths of both.
Amanda Murphy, an analyst with equity investment firm William Blair in Chicago, Illinois, says that the market for Complete Genomics’ expertise is poised to grow. Most human sequencing has so far focused on exomes, the roughly 1.5% of the genome that codes for protein. But the recent ENCODE project3 and other research is revealing functions of non-protein-coding regions. “I think the world is moving towards whole-genome sequencing,” she says. “The demand just wasn’t there fast enough” for Complete Genomics to hang in on its own.
Demand for services is increasing as well. Many researchers want to use sequencing data but do not want to invest in the expertise and instruments needed to do the sequencing themselves. Even large academic sequencing centres often have more sequencing projects than capacity.
A major drawback of Complete Genomics’ technology is that it takes 2 to 3 months to produce a sequence, which discourages medical uses. But the platform is expected to get faster. Peter van der Spek, head of bioinformatics at Erasmus Medical Center in Rotterdam, the Netherlands, worries that innovation will stall under BGI, although a spokesman for BGI says that Complete Genomics’ current scientific staff is expected to stay with the company and that development would continue.
Either way, the community is better off with options, says Richard Gibbs, director of the Human Genome Sequencing Center at Baylor College of Medicine in Houston, Texas. “If we just had one test and one machine and one process, things would happen, but not as fast.”
- Journal name:
- Date published: