Later versions of the HeliScope would have a throughput of more than 1 billion bases per hour, or 10 billion bases per day, according to Lapidus.
That throughput could be achieved using flow cells with higher template densities, better assay conditions, and new software, according to Lombardi. The system has been designed and engineered to be able to achieve this higher output of “alignable” sequences, he wrote in his e-mail.
Lapidus did not mention the system’s read length, but in response to an investor’s question, he said that “if you can deliver 25 bases on each end of a paired-end ready, you are good,” adding that even shorter reads are satisfactory for tasks like gene expression analysis. “Read length is application dependent,” he said.
According to Lombardi, the system’s current mean read length is 27 bases.
Helicos, Lapidus pointed out, sees greater market opportunity in medical re-sequencing and other areas that do not require long reads, rather than in de novo sequencing, where longer reads are advantageous.
454’s new instrument will provide read lengths of 200 bases or greater, Solexa’s 1G Genetic Analyzer currently creates 25-base reads, and ABI is hoping its first commercial units will deliver 25- to 50-base reads, depending on the template library.
Lapidus also touted the HeliScope’s ability to contain cost per base and hour, claiming that the company will be “quite strong” on this.
“We will sequence one human genome every 10 days at a cost of less than $10,000 per genome,” he said.
That number is “the best estimate we have at this point” for sequencing a human genome at 10x coverage with the first generation flow cell, assay, software, and bioinformatics, according to Lombardi.
In addition, Lapidus stressed that Helicos’ single molecule sequencing chemistry will greatly simplify sample preparation since it does not involve making libraries or clones. “We shear and sequence,” he said.
The company also has a strategy for paired-end reads that does not require libraries, he said: After reading a stretch of sequence, scientists add a number of “dark,” or unlabeled bases, then resume reading the template with labeled bases.
The immediate market for the HeliScope will be academic and pharmaceutical research, Lapidus said, but “the largest application may well be in diagnostics,” he added.
But the company has also been working with Floyd Romesberg at the Scripps Research Institute on improving DNA polymerase, and “will announce a couple of collaborations soon,” Lapidus said last week.
Helicos has raised a total of $67 million in funding so far, $27 million in early 2004 and $40 million earlier this year. Investors in both rounds included Flagship Ventures, Atlas Venture, Highland Capital, MPM Capital, and Versant Ventures. The company anticipates another financing round in late 2007 or early 2008 to help commercialize the HeliScope, according to Lapidus.
The company has licensed “a fair bit” of intellectual property, he said, including single-molecule sequencing technology invented by Steve Quake at Caltech and published in PNAS in 2003, as well as from other inventors.
Though the HeliScope is likely to face some tough competition, Lapidus is confident his company’s single-molecule sequencing platform will stand out from the crowd. “Our data will ultimately speak for itself,” he said.