Skip to main content
Premium Trial:

Request an Annual Quote

Helicos Still Validating Performance; Initial Specs Match Next-Gen Competition

The initial version of Helicos BioSciences’ single-molecule sequencing system promises to generate DNA sequence data at a comparable price, accuracy, and throughput to competing systems, according to specifications the company disclosed last week.
Helicos is still testing production units of its HeliScope platform in house but said it is “on track” to ship the instruments to customers, a goal it had originally planned to meet by the end of last year.
Speaking to investors at the JPMorgan Healthcare Conference in San Francisco last week, Helicos president and COO Steve Lombardi said that for sequencing applications, the first HeliScope version will generate 7.5 gigabases of data per two-week run, or about half a gigabase per day, for $18,000 in reagents. These reagent costs translate to $2,400 per gigabase. He also showed that the error rate, which is independent of the read length, is between 1 percent and 1.5 percent.
Lombardi did not mention similar specifications for gene-expression analysis, the other main application the company sees for its system. In the past, it has said that the system will initially produce DNA sequence data, an application in which each strand is sequenced twice to increase accuracy, at a rate of 25 megabases per hour, and gene-expression data at a rate of 90 megabases per hour.
Given the specifications for sequencing Lombardi disclosed last week, the HeliScope promises to have a comparable sequencing performance to other next-generation sequencers such as Illumina’s Genome Analyzer or Applied Biosystems’ SOLiD system.
According to the most recent specifications from vendors of these systems (see In Sequence 1/1/2008), Illlumina’s Genome Analyzer produces approximately 2.6 gigabases of paired-end sequence data in a 6-day run, or 0.4 gigabases per day, with a per-base read accuracy greater than 98.5 percent, which corresponds to an error rate less than 1.5 percent. As of May 2007, Illumina cited a reagent cost of about $3,000 per gigabase.
ABI’s SOLiD generates up to 4 gigabases of paired-end data in a run that takes up to 10 days, also translating to 0.4 gigabases per day, with an accuracy of 99.94 percent after error correction from two-base encoding, or an error rate of 0.06 percent. ABI put the cost of sequencing at $2,300 per gigabase.
With a list price of $1.35 million (see In Sequence 11/6/2007), the HeliScope is more than twice as expensive as Illumina’s and ABI’s platforms, which list for around $430,000 and $591,000, respectively.
But Helicos said it believes that it will be able to improve the performance of its platform significantly in the future, citing a possible imaging rate of 1.3 gigabases per hour, without having to change the instrument itself.
“We believe we can get two orders of magnitude better performance,” Lombardi said last week. “And the instrument we are shipping on day one has the ability to do that today. No upgrades required, just better chemistry, available through the next kit.” He did not give a timeline for these improvements.
He also pointed out that the tool’s sample preparation does not involve an amplification step like the other next-gen sequencing systems, since the HeliScope reads single molecules.
“It’s one person, with a pipetter, working for a few hours,” Lombardi said. Illumina cites an 11-hour sample prep time for the Genome Analyzer, while ABI specifies about a week for the SOLiD system.
The system also offers real-time image processing, which ABI has said its SOLiD system can also do.
Lombardi also stressed that the HeliScope is able to multiplex, as it is capable of running samples in 50 independent channels on two flow cells. Read length and accuracy still vary significantly between channels, he showed, but “it’s a big part of our testing right now to get this variability out.”
Illumina’s Genome Analyzer allows users to run samples in eight separate channels on a single flow cell. Like the HeliScope, ABI’s SOLiD can run two flow cells in parallel, each of which can take up to eight samples.

“We believe we can get two orders of magnitude better performance.”

He also mentioned examples of proof-of-principle studies Helicos has performed on the system internally. For example, in a chemogenomics application, company scientists have measured expression signatures of 89 genes in 3,000 samples, following treatment with two chemicals. The expression patterns they observed resulted from the chemicals and were not due to technical issues, he said.
Measuring gene signatures like these, Lombardi said, could tell pharmaceutical companies “whether a drug is responding one way or the other.”
Using “traditional technology,” he claimed, an experiment like this would cost “a couple of hundred dollars” while it cost 20 cents on the HeliScope. “So it has the ability to do what pharma wants to do, but at a cost that allows them to deploy those signatures — be they toxicology, be they biomarkers, be they pathways like this — to be able to use them in downstream drug development.”
In an example of a medical resequencing application, Helicos researchers sequenced 29 exons in four candidate oncology genes with a 1.4-percent per-base error rate.
“This is a model system, we know what the reference is, we want to prove that the chemistry can find the sequences of importance,” he said. The average coverage was 50-fold but varied significantly between bases. “At that throughput, you can still find the key mutations,” Lombardi said. Using the company’s virtual terminator chemistry, the scientists were also able to sequence through homopolymer regions, he added. He did not mention the cost of that study.
Lombardi also mentioned a new collaboration between Helicos and Victor Ambros’ group at the University of Massachusetts Medical School to develop an assay to characterize microRNAs and discover new non-coding RNAs. He said Helicos CSO Patrice Milos will present data from this collaboration at the Advances in Genome Biology and Technology meeting on Marco Island next month.
In terms of commercialization, Lombardi said the company is “on track to ship our first production units” but acknowledged that the final testing is time-consuming. He did not mention any sales or a timeline for shipments.
Last fall, Helicos said it expected to obtain its first orders and ship instruments by the end of 2007 “or soon thereafter” (see In Sequence 11/6/2007).
Helicos, Lombardi said last week, is working with a total of nine instruments on its manufacturing floor, two of which are “in [the] final stage of production readiness.” Each “verification and validation” test run on these instruments takes about three to five days and generates terabytes of data, he said.
Extensive testing is necessary “so when we do have problems out in the field — and we will — we know how to diagnose and fix” them, Lombardi said.
Lombardi did not say whether Helicos has sold instruments yet but said “we have sent out many contracts, we have got lots of interested parties, many more prospects than we have instruments on the manufacturing floor.” Prospective customers include pharmaceutical companies, academic health centers, and genome centers, he said.
But he said the company is confident that sales will happen soon. “What we are hearing, after [potential customers] see this data, is, ‘We can’t afford not to buy one of these.’”
Helicos hopes to capture growth margins of up to 60 percent from the instrument, and between 70 and 80 percent from reagents and disposables “when we hit scale,” he said.

The Scan

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.

Coronary Artery Disease Risk Loci, Candidate Genes Identified in GWAS Meta-Analysis

A GWAS in Nature Genetics of nearly 1.4 million coronary artery disease cases and controls focused in on more than 200 candidate causal genes, including the cell motility-related myosin gene MYO9B.

Multiple Sclerosis Contributors Found in Proteome-Wide Association Study

With a combination of genome-wide association and brain proteome data, researchers in the Annals of Clinical and Translational Neurology tracked down dozens of potential multiple sclerosis risk proteins.

Quality Improvement Study Compares Molecular Tumor Boards, Central Consensus Recommendations

With 50 simulated cancer cases, researchers in JAMA Network Open compared molecular tumor board recommendations with central consensus plans at a dozen centers in Japan.