This article was originally published Oct. 24.
BOSTON – Illumina plans to launch an upgrade for the HiSeq early next year that will increase the system's output to 1 terabase of data per six-day run. The company will make 2x250 base pair reads for the HiSeq 2500 available in mid-2014, later than originally planned.
During a company workshop at the American Society of Human Genetics annual meeting in Boston last week, Illumina said that the HiSeq upgrade will be available in the second quarter of 2014 and will only involve changes to the system's firmware and software but not its hardware.
As a result, the system will be able to generate 1 terabase of data in a six-day run in high-output mode, using 2x125 base paired-end reads and 4 billion clusters. The improved output is made possible by an increase in the number of clusters per flow cell and by longer reads.
Jeremy Preston, Illumina's director of product marketing for systems and consumables, showed data from a recent internal Illumina run that generated 1.035 terabases of data and 4.14 billion reads in six days, with 88 percent of bases having greater than Q30 quality.
Customers could use the upgraded HiSeq to sequence 10 whole human genomes at 30x coverage in a single run, or 150 human exomes, or 80 whole transcriptomes with 50 million reads per sample, he said. The price per data point will be "significantly reduced" compared to the current price.
All factory-built 2500 and 1500 systems are eligible, as well as 2000 and 1000 systems that were shipped in 2013. Pricing for the upgrade, which he said is "minor," has not been finalized yet.
Preston told In Sequence that no customer has the upgrade installed at the moment, but Illumina will soon make it available to a limited number of early-access customers, as it usually does with new technology.
Illumina had already mentioned a 1-terabase HiSeq run two years ago at the Advances in Genome Biology and Technology conference (IS 2/22/2011), but at the time that run took 14 days, albeit with longer reads of 2x150 base pairs. Preston said Illumina decided that run time was too long, so it waited until it could deliver a six-day run.
Based on customer feedback, the company prioritized commercialization of the 1-terabase run, he said, and as a consequence, it pushed back the launch of 2x250 base pair reads for the HiSeq 2500 in rapid run mode, originally planned for the second half of this year (IS 2/26/2013), to mid-2014.
Internally, Illumina recently obtained a 147-gigabase single flow cell run with 2x250 base pairs on the HiSeq 2500 in 60 hours, he said, generating almost 300 million reads, with 88 percent of bases greater than Q30.
Regarding the MiSeq, Preston said that Illumina's best internal run to date using the new chemistry that was launched in August generated 19 gigabases of data, with 91 percent of the bases greater than Q30, using 2x300 base pair reads. Early users of the new chemistry have achieved 16.4 gigabases of data per run on average, with 86 percent of bases at Q30 or better.
Based on an analysis by Illumina of the number of data submissions to the National Center for Biotechnology Information's Sequence Read Archive, more than 85 percent of desktop sequencer data is currently generated on the MiSeq, he said.