By Monica Heger
This story was originally published Aug. 4.
Pacific Biosciences said this week that installations of 16 PacBio RS single-molecule systems generated the lion's share of its $10.6 million in revenue for the second quarter of 2011, which beat analyst estimates of $4.1 million.
The installations, which included 11 beta systems that were fully upgraded during the quarter as well as five new systems, contributed $10.1 million to PacBio's quarterly revenue. The company posted $629,000 in revenue in the second quarter of 2010.
PacBio also booked seven new orders during the quarter, bringing its backlog to 35 systems as of June 30, 2011, which it expects to book as $22 million in revenue by the end of the year. In line with its previous estimate, it expects to generate around $35 million in revenue for 2011.
"For our first quarter of selling a product, things went really smoothly," said PacBio's CEO Hugh Martin in a conference call with investors to discuss the company's second-quarter earnings.
The company is also planning an upgrade of its chemistry and consumables for the fourth quarter of this year, which it has dubbed C2. The improved chemistry will increase read lengths, improve accuracy, and double the throughput of the system.
With the C2 upgrade, read lengths will average 2,700 base pairs, with 5 percent of reads achieving lengths of 5,100 base pairs or higher. Single-pass raw read accuracy will increase from 85 percent to 87 percent. For double-pass circular consensus sequencing, accuracy can be brought up to 93 percent using a 1,350-base pair insert, and to 99 percent using a 500-base pair insert. Throughput will also be doubled to 90 megabases per SMRT cell.
The improvements will enable customers to "take on the entire spectrum of sequencing projects," including de novo assembly without using other techniques, said Martin.
The improvements to the chemistry will make obsolete the company's "strobe sequencing" feature, in which the machine sequences long stretches of DNA in bursts. While the company initially promoted strobe sequencing as a way to improve assembly by joining contigs, early-access customers reported in May that the feature still needed work (IS 5/17/2011). Now, with longer read lengths, the company is no longer focusing on that application.
"As we are getting the read length up, customers are reluctant to invest time in strobe sequencing. The longer the read length, the less you need strobe," Martin said. "It's not a major thrust for us" anymore.
Slow on the Uptake
Despite these improvements and strong initial sales, some investors remained concerned about the future uptake of the system. Following the earnings call, investment firm Oppenheimer lowered revenue estimates for the company to $66 million in 2012, $119 million in 2013, and $165 million in 2014 from, $87 million, $156 million, and $230 million, respectively.
"We worry about PacBio RS placements due to lower-than-industry accuracy, long lead times to placements, and high costs in the face of funding environment challenges," analyst David Ferreiro wrote.
The seven new orders PacBio booked in the second quarter represent only one more than the first quarter.
Meantime, a William Blair analyst said that while the backlog growth this quarter was a "touch lighter than expected," it has not changed its 2012 or 2013 revenue estimates of $108 million and $181 million, respectively.
Investors reacted strongly to the backlog concerns, causing the company's shares to fall 34 percent from a closing price of $9.90 on the day it reported its earnings to a closing price of $6.50 the following day.
[ pagebreak ]
PacBio's Martin said that aside from the funding uncertainty in the industry as a whole, there were two likely factors affecting new orders, but he does "not foresee them hindering long-term prospects."
The first issue was that the beta systems had some initial variability problems, which, while not surprising, likely kept customers "from truly assessing the instrument," Martin said. Additionally, the initial raw read accuracy of 85 percent limited users. But, "we believe this issue will be eliminated" with the launch of the C2 chemistry, which will increase raw read accuracy, he said.
Martin added that the recent publication by PacBio researchers and colleagues in the New England Journal of Medicine of the genome sequence of strain of Escherichia coli responsible for the outbreak in Germany would help spur uptake of the system (IS 8/2/2011).
The publication, which served as the first public example of the company's C2 chemistry, demonstrated the ability to use PacBio's long read lengths as well as the accuracy of the circular consensus sequencing to generate a de novo assembly based only on PacBio data, Martin said.
PacBio's 'Sweet Spot'
Another good sign, said Martin, is that PacBio RS customers such as the Department of Energy's Joint Genome Institute, the Broad Institute, and the Sanger Institute are "running them around the clock."
This is particularly encouraging because those institutions are typically multi-unit buyers, Martin said. Although the company has not yet received multiple orders from an existing customer, the fact that machines are "fully booked" at customers who typically have large installations of systems is promising. Users are "starting to find a sweet spot of where [the RS] will be useful," he said.
Of customers with the machine, 90 percent are using it for targeted resequencing applications, 75 percent for either de novo sequence and assembly or hybrid assembly with sequence data from other platforms, 60 percent for pathogen identification, and 20 percent for transcript analysis.
Martin cited some early projects for which its customers are using the machine, including a recently announced targeted resequencing project of cancer genes in a clinical trial at the Ontario Institute of Cancer Research (CSN 8/3/2011), as well as a project being done at Mt. Sinai Hospital in New York, where chief scientific officer Eric Schadt has a joint appointment, to monitor the emergence of drug-resistant bacteria such as Staphylococcus aureus and Clostridium difficile.
One emerging application is to use the PacBio RS to correct errors in previously assembled genomes, including assemblies that have been filed in GenBank. "The long read lengths enable [users] to go back in and create platinum versions of assemblies," said Martin.
Despite the fact that some of PacBio's applications — such as targeted resequencing and rapid pathogen identification — overlap with those of some of the newer desktop sequencers, like the Ion Torrent PGM and Illumina MiSeq, Martin said he did not see those platforms as being the company's competition.
"We're really in a very different space," he said, and will compete more with capillary electrophoresis machines or Roche's 454 GS machines.
In line with previous estimates, PacBio still expects to generate $35 million in revenue for 2011.
It reduced its second-quarter net loss to $22.5 million from $32.7 million in the year-ago period.
Research and development expenses fell by 28 percent to $19.5 million from $27.1 million in the second quarter of 2010. Sales, general, and administrative expenses, meantime, rose 77 percent to $11 million from $6.2 million in the second quarter of last year.
PacBio ended the quarter with $216.6 million in cash and investments.
Have topics you'd like to see covered by In Sequence? Contact the editor at mheger [at] genomeweb [.] com.