By Julia Karow
This article, originally published March 9, has been updated with information from Complete Genomics' fourth-quarter earnings call.
Complete Genomics said last week that it has signed orders for more than 1,400 human genomes so far this year, including a previously announced 615-genome order from the Institute of Systems Biology but excluding an existing multi-year contract with the National Cancer Institute that may add another 1,128 genomes.
The company has grown its customer base to more than 45, cut the median turnaround time for its service to 73 days, and expects to book revenue for more than 500 human genomes during the first quarter.
For the fourth quarter of 2010, the firm posted $3.8 million in revenue from more than 300 genomes, along with a net loss of $10.3 million.
By the end of this year, Complete Genomics plans to cut its price per genome from a current $9,500 for small orders to less than $5,000. In addition, it is working on a number of technological upgrades that will increase the throughput of its sequencers by tenfold and improve the accuracy of its genomes by a hundredfold in 2012.
As of early February, the company had sequenced more than 1,000 genomes for customers in total at high coverage. At the moment, its Mountain View, Calif.-based facility has the capacity to sequence 400 human genomes per month, but CEO Cliff Reid said that the firm expects to ship just 500 human genomes in the first quarter.
Last year, the project turnaround time was between 90 and 120 days, he said, but the median turnaround for the last three months was just 73 days — 10 days shorter than earlier this year — which he said is "comparable" to competitors. The company plans to reduce it to 60 days by the end of the year.
Besides price, turnaround time is an "important variable," Reid said, since most of the company's current customers are researchers who want to publish their results quickly.
Complete Genomics has grown its customer base to more than 45, up from 35 in December (IS 12/14/2010). As of the end of 2010, more than half the customers who received and analyzed data had placed a second order, Reid said. While many first-time customers still start out with a small pilot project, the company has seen a "handful of examples" of new customers placing larger initial orders, he added.
In order to grow its business, the firm has increased its sales team from two at the beginning of 2010 to 12 by the end of the year, reflecting a "small but growing presence in the US and in Western Europe," Reid said. This year, the company plans to double its sales force.
Also, more than 350 researchers have downloaded a dataset of 60 human genomes that Complete sequenced with its technology and released last month (IS 2/8/2011). "We believe access to this data will be a powerful factor in securing new customers," Reid said.
Another factor in gaining more customers will be price reductions. The company currently offers human genomes to customers for about $10,000, which Reid said is "well below what researchers can achieve on their own."
[ pagebreak ]
By the end of this year, the firm plans to offer its "basic" human genome sequencing service for less than $5,000 and, as a result, expects that the service "will be adopted by researchers who have been limited to analyzing exomes because of their budget constraints," Reid said.
The company is considering whether to charge more for "premium services," such as highly accurate cancer genomes or analysis and storage solutions in the cloud. Regarding the latter, the firm announced a partnership with DNAnexus this week, under which its customers can store and visualize human genome sequencing data using DNAnexus' cloud-based platform (see other story, this issue).
Complete Genomics currently uses Amazon web services to deliver genome data to customers. Most customers don't download the data because they don't have enough bandwidth, Reid noted, but have it copied by Amazon onto a disk drive and shipped to them by FedEx.
"We think the biggest driver for uptake is still the price," Reid said, and the next goal beyond the $5,000 mark will be to break the $3,000 price barrier, which "takes us into the price of clinical sequencing."
Customers are unlikely to delay orders because of pending price cuts, he said, because once they have funding in hand, they are "under significant time pressure to conduct their projects and turn that into medical results or publications."
In order to "make the economics of the $5,000 genome work," he said, the company needs to further drive down its internal costs through process improvements across its entire service.
This will include technological improvements to its sequencing platform. Right now, each of its proprietary instruments generates more than 150 gigabases per day, equivalent to about one human genome.
Through improvements in array density, fluidics efficiency, and camera speed, a new generation of instruments will be able to produce data for more than 10 genomes per day. This will cut the instrument amortization and utilization cost per genome from about $1,000 to $100. He did not say what sort of reduction the company expects in consumables costs, which are currently "substantially below" $2,000 per genome.
Reid said the company aims to have "working prototypes" of the new sequencers this year, and have them "ready for commercial use" in 2012.
Through a combination of process improvements and additional sequencers, Complete's sequencing capacity is expected to grow to between 800 and 1,200 genomes per month by the end of the year, and the new instrument generation will "greatly enhance our ability to scale our operations at our current Mountain View facility," he said.
The current throughput of one genome per day and instrument relies on ordered nanoarrays with 3 billion spots that each bind a DNA "nanoball." The spots on the array are then lined up with the pixels of the camera. Reid noted that this feature "gives us enormous camera efficiency," which is the rate-limiting step for the speed of any camera-based DNA sequencer.
At the moment, each DNA spot is aligned with two camera pixels, but by the middle of the year, the company will make nanoarrays that utilize only one pixel per spot. This technology is "heavily protected by a patent portfolio, and nobody is going to be able to practice these kinds of arrays without coming to us," Reid said.
The company is also further reducing the amount of DNA required as input. It recently cut that amount from 15 micrograms to 7.5 micrograms, Reid said, and will continue to do so as it improves the efficiency of its protocols. Longer term, non-PCR amplification technologies with little bias, currently in development in the industry, will reduce the amount of DNA needed to picograms, he predicted.
Reid also provided some information about the company's "long fragment read technology" for haplotyping, which it plans to introduce to early-access customers in the fourth quarter and make widely accessible next year.
He said the patent-protected method, which separates the two parental chromosomes, involves an extra step each in sample prep and informatics. It relies on hierarchical fragmentation of the genome, which is first broken into 100-kilobase fragments. These fragments are then barcoded and broken up again for sequencing. The sequence reads are reassembled into 100-kilobase fragments. The method provides "the informatic power of 100,000-base read lengths, which no one can do without paying the cost in accuracy and expense of doing super-long read lengths," Reid said.
[ pagebreak ]
A "wonderful added benefit" of the method is that it cuts the sequencing error rate from 10-5 to 10-7, or from 30,000 errors per genome to 300 errors per genome, which he said is good enough for clinical applications.
Reid cited UBS estimates that pegged the current market for human whole-genome sequencing at around $50 million, for research only, with the potential to grow to about $1.45 billion by 2015, consisting of $900 million for research sequencing and $550 for clinical sequencing.
He said that the company sees three major emerging markets for its service: cancer sequencing, clinical trials, and consumer genomics. Since each of these needs to be "highly standardized," he said they will be provided best not by genome centers who deal with different types of projects and organisms, but by an organization like Complete Genomics, which "is a factory that can do extremely high-throughput, high-quality, and low-cost standardized complete human genome sequencing."
In the future, he said, Complete Genomics will continue to do "more of the same" in terms of improving its technology, cost, genome quality, and scale. "This is a scale game," Reid said, noting that the company's success will depend on how many genomes it can ship for the material, equipment, labor, and overhead invested.
As previously mentioned, the company also plans to build additional sequencing facilities around the world in order to access new markets, in particular in Asia. Reid did not provide more details but volunteered that he had just returned from a week-long trip to China.
During Q4 of 2010, Complete Genomics booked $3.8 million in revenues, up from $600,000 during the year-ago quarter, when it had not yet begun full commercial operations.
Research and development expenses for the quarter totaled $5.6 million, down from $6.3 million during the year-ago quarter.
The company's net loss for Q4 amounted to $10.3 million, similar to last year's Q4 net loss of $10.4 million.
For all of 2010, Complete Genomics recorded $9.4 million in revenue, compared to $600,000 in 2009.
R&D expenses dipped to $21.7 million for the year from $22.4 million in 2009.
The firm's net loss in 2010 was $57.7 million, up from $35.9 million in 2009.
As of Dec. 31, Complete Genomics had $68.9 million in cash and cash equivalents.