When Helicos BioSciences’ single-molecule sequencing system hits the market, it will have a list price of $1.35 million — more than double the price of the most expensive next-generation sequencer on the market to date, company officials disclosed last week.
Along with the pricing information, Helicos released some technical specifications for its Helicos Genetic Analysis system and an update on its commercialization plans. However, the company has yet to publish some key specs for its system, such as the sequencing error rate and cost per base, as well as other data that experts say is needed to compare its performance with that of other next-generation sequencing platforms.
During the company’s third-quarter earnings call last week, president and chief operating officer Steve Lombardi said that Helicos is currently generating data that will allow it to obtain its first orders and ship instruments “by the end of this year or soon thereafter.” Previously, the company said it was planning to ship its first instruments by year’s end.
Helicos declined to provide further information on the system for this article.
The $1.35 million list price of the Helicos Genetic Analysis system includes the HeliScope — the chemistry and imaging system — as well as the HeliScope Analysis Engine, a data-processing and data-storage unit with 28 terabytes of disk space that provides “near real-time processing of the data.” In addition, the system comes with a sample loader, software, startup reagents, installation, and a one-year warranty.
For comparison, ABI’s SOLiD system, until now the most expensive next-gen sequencer, lists for around $600,000, which includes a computer system with more than 11 terabytes of storage.
Two different reagent kits for the Helicos system — one for single-pass gene expression applications, or “genomic signature sequencing,” the other for sequencing with high accuracy, where each strand is sequenced twice — cost $18,000 each, but it is unclear how much data they yield. As a result, the cost per base on the system is unknown.
Paired-end reads, which are still in R&D, will not be available initially, and the company is just beginning to test its “virtual terminators,” which promise to reduce homopolymer errors.
The company did not provide details on sequencing error rates, a metric that many potential customers will want to know before deciding whether to purchase the system.
“I think right now, the single most important information people want to know about Helicos to make a decision is: ‘What is the raw base error rate?’” Yuan Gao, an assistant professor at Virginia Commonwealth University, told In Sequence by e-mail. Gao acquired an Illumina Genome Analyzer earlier this year and has been in contact with Helicos about its system in the past.
Helicos notes on its website that the average per-base accuracy is unaffected by the read length, but it does not say what the accuracy actually is. During the call, Lombardi said that for the dual-pass sequencing mode, where each base is sequenced twice, “our goals remain to have [errors] of less than 1 percent,” but he did not say if that error rate assumes the use of virtual terminators.
For comparison, Illumina specifies a raw accuracy of greater than 98.5 percent, equivalent to an error rate of less than 1.5 percent, and ABI states a raw base accuracy of more than 99.94 percent, or an error rate of less than 0.06 percent, after error correction with its two-base encoding scheme.
Helicos also did not say how much data per run the instrument will deliver, or how long a run will take, but reiterated its aims for an initial throughput of 25 million “usable bases” per hour for sequencing applications, which require two-pass sequencing, and 90 million such bases per hour for one-pass gene expression applications.
“Those rates remain our goal, and we are showing progress,” said Lombardi during last week’s call. The 25 megabase-per-hour rate takes into account the loss of strands over several consecutive sequencing cycles, and only includes reads above 25 bases, Lombardi and CEO Stan Lapidus said during the call. The company believes that its system can ultimately achieve rates of up to a billion bases per hour.
The initial rates would scale to between 600 megabases and 2.2 gigabases per day. For comparison, Illumina’s Genome Analyzer produces up to 400 megabases per day of “filtered, high quality data”, according to the company, and ABI specifies up to 4 gigabases of “mappable” data from a mate-pair library in a 10-day run.
Helicos explains on its website that almost half its reads have a length of 30 bases or longer — unfiltered for errors — after 28 full sequencing cycles, or “quads.” Illumina currently provides 35-base reads as a standard, and ABI’s SOLiD system generates 35-base reads for single reads. However, Helicos did not provide details on the read length distribution after the second pass of sequencing for high-accuracy applications.
The company currently has six “commercial-grade” instruments in house, which are “in various stages of assembly and testing,” Lombardi told investors last week. These are currently being tested for “production readiness,” he said.
Company researchers are now using the system to generate gene expression data in “biological experiments,” he said, and the company’s genomic sciences group plans to test sequencing applications in “large-scale collaborations with key customers.”
“I think right now, the single most important information people want to know about Helicos to make a decision is: ‘What is the raw base error rate?’”
The company’s virtual terminators, designed to eliminate homopolymer errors and first presented at a meeting in February (see In Sequence 2/13/2007) are now in “process development,” Lombardi said. “Those molecules now have been synthesized to a point where we have transferred them back to the genomic sciences group to begin doing work on generating collaborations and good data,” he said.
Paired-end reads are also still in development. “That work is still in our research program that’s partially funded by our NIH grant, and we continue to work on it,” Lombardi said. The three-year NIH grant, awarded in 2006, is worth $2 million.
Data for a ‘Skeptical’ Market
The company has identified 30 to 40 potential customers that have “the charter, the funding, the samples, and the infrastructure” to adopt the HeliScope, he said, among them not only large genome centers but also academic health centers. “We are getting a tremendous amount of interest from them,” he said of the latter.
Helicos’ sales team has already talked to prospective users so it can obtain orders soon, but will have to prove first what the system is capable of. “During our ongoing verification and validation process of testing HeliScopes, we expect to generate the data required as proof of performance to obtain these orders,” Lombardi said.
He acknowledged that this process might not be easy. “These customers are skeptical in nature and absolutely want the best that they can get,” he said. “Our best estimates have us acquiring orders and shipping our first production HeliScopes by the end of the year or soon thereafter.”
For the third quarter, Helicos reported $230,000 in revenues from its NIH grant, up from no revenues during the same period last year.
The company had $7.2 million in R&D expenses for the quarter, up from $4.2 million in the year-ago quarter.
Helicos’ net loss for the quarter amounted to $10 million, up from $5.9 million during last year’s third quarter.
As of Sept. 30, the company had $54.9 million in cash and cash equivalents. Helicos still expects to end the year with at least $40 million of cash in the bank.
Following the company’s earnings call last week, UBS downgraded Helicos from “neutral” to “sell,” while other research firms maintained their previous ratings.
"While the anticipated specifications on the HeliScope’s performance are unchanged, we have yet to see hard data demonstrating the instrument can routinely and robustly deliver the throughput and accuracy," UBS analyst Derik De Bruin and colleagues said in a note to investors, noting that Illumina and ABI have recently improved their systems.
The UBS analysts also criticized the instrument’s list price as “prohibitive given the initial performance is likely to be on par with the competition.”
They also voiced concerns about the potential delay of first shipments until next year. “[W]e begin to wonder whether product development is on track. While we are aware that the R&D process is inherently an unpredictable one, we stressed during our initiation [of coverage] that we believe time is of the essence for Helicos, and the margin for error is slim given the fierce competition for scant R&D dollars.”