This article was originally published Oct. 26.
Pacific Biosciences booked four new systems in the third quarter of 2012, an increase from just one in the previous quarter, but down from six in the year-ago quarter. The Q3 orders bring its backlog to five systems.
Additionally, the company is planning to launch new chemistry in the fourth quarter that it says will increase average read lengths to 5,000 bases with five percent of reads above 13,000 bases. PacBio's current C2 chemistry yields average reads of 3,000 bases with five percent above 8,000 bases.
"Long reads are especially valuable for de novo assembly applications and targeted sequencing of large repeat regions and regions with complex structural variation," CEO Mike Hunkapiller said during a conference call discussing the company's 2012 third-quarter earnings.
Nevertheless, the company faced steep revenue declines this quarter — down 73 percent to $2.8 million from $10.5 million in the year-ago quarter and missing analysts average estimates of $3.1 million.
The company did not recognize any revenue from sales of its PacBio RS sequencer during the quarter. It booked just one system in the previous quarter, which was still in the process of being installed in the third quarter, Hunkapiller said.
He added that this was to be expected, as the company was not focused on booking instruments in the first half of the year, but that bookings should pick up in the second half of 2012.
"We made progress on this goal in the third quarter with bookings for four new systems," Hunkapiller said, which the company attributes to "the progress we've made in reliability and performance and focusing on making our customers successful."
The company expects to recognize instrument revenue in the next quarter, as it installs some of the five systems currently in its backlog, Ben Gong, vice president of finance and treasurer, said during the call.
As the company said in previous earnings calls this year, it focused on the first half of the year on making improvements to the PacBio RS system.
Hunkapiller said he is "pleased" with the progress the company has made on this front, noting that "system reliability has come a long way" since the beginning of the year. Customers now have the C2 chemistry upgrade, which has significantly improved "system uptime," he said.
Additionally, two products launched in recent months — the MagBead sample-loading station, which increases the number of long reads and lowers input requirements; and new software to detect base modifications — have made a big impact on instrument utilization, he added.
Going forward, the company expects base modification analysis to be a key application of the PacBio RS.
Last quarter, the company said it had developed new chemistry that it was testing in house (IS 7/31/2012), and this week, Hunkapiller said that the new chemistry would be launched in the fourth quarter of this year.
The chemistry increases average read lengths from 3,000 to 5,000 bases with the longest reads over 13,000 bases.
The increase in read lengths and subsequent increases in throughput will enable new applications, Hunkapiller added. For example, he said the company has already received positive feedback from an early-access customer that is using the chemistry to sequence the rice genome, which is "replete with repetitive elements and duplications."
Another of the company's goals for 2012 was to focus on its existing customer base, betting that happy customers would drive adoption of the system (IS 5/8/2012).
The company recently held a user meeting that drew 78 attendees from 25 different customer sites, Hunkapiller said, citing two customer presentations that highlight the RS system's capabilities.
In one project, researchers from the Center for Genomics Sciences at the Allegheny-Singer Research Institute sequenced and assembled bacterial strains from both human and plant pathogens, using only PacBio sequencing and the Celera assembler, Hunkapiller said.
As part of a project for the US Department of Agriculture, the group sequenced and assembled 20 strains of a plant pathogen that causes diseases in more than 100 plant species, using four SMRT cells per genome. Additionally, the group just received funding from the National Institutes of Health for a project to sequence around 1,700 influenza strains in order to understand pathogen evolution. For this, the team will use three SMRT cells per genome.
A second group, at the Human Genome Sequencing Center at Baylor College of Medicine, has been using the RS to improve assembly of primate genomes, Hunkapiller said. The Baylor group is studying the genome of the sooty mangabey monkey, which is naturally resistant to AIDS, and comparing it to the macaque genome, which is extremely susceptible to the disease, in order to identify the genetic pathways associated with resistance and susceptibility.
While a draft genome for the sooty mangabey exists, Hunkapiller said that the genome, which had been sequenced to 100-fold coverage using Illumina technology, contained 186,000 gaps and had an N50 contig size of 35,000 bases.
Sequencing on the PacBio reduced the gaps by 65 percent and tripled the length of the contigs, he said. The Baylor team plans to publish the software that it used for the assembly in PLoS One, he added.
Product revenue was $1.3 million, down from $9.8 million in the year-ago quarter, while service and other revenue increased to $1.3 million from $535,000 in the comparable period of 2011.
Net loss narrowed to $22.7 million from $29.2 million in the third quarter of 2011.
R&D costs fell 37 percent to $12.6 million from $20 million, while SG&A spending declined 21 percent to $10.1 million from $12.8 million.
Looking ahead, Gong said that he expects fourth-quarter revenues to "increase sequentially," with expenses remaining stable. Additionally, he said the company is targeting less than $20 million of cash use per quarter, so it should finish the year with around $100 million in cash, not including any additional capital the company may raise.
As of Sept. 30, 2012, cash and investments totaled $119.4 million.