A $250 disposable chip lies at the heart of Life Technologies' new DNA sequencer. Credit: Ion Torrent

A benchtop machine that reads DNA using semiconductor technology could help make gene sequencing affordable to labs around the world. The machine's maker, Life Technologies in Carlsbad, California, announced today that it has begun to ship orders of its Ion Personal Genome Machine (PGM) sequencers, which sell for US$49,500 each — a price that is affordable to many labs. And in a bid to improve the technology, the company is launching a multi-million-dollar crowd-sourcing competition.

The Ion PGM, developed by Ion Torrent — a start-up company based in Guilford, Connecticut, that was bought by Life Technologies in August — is the first of a new wave of sequencing machines to hit the market. Most current sequencing technologies label nucleotides with dyes, which must then be optically read as sequences, but the Ion PGM functions by directly detecting electrical signals on a disposable chip costing $250.

"This really is different than all the machines we've had before," says Jonathan Rothberg, founder and CEO of Ion Torrent. "This isn't optics — it's just a computer."

Most sequencing machines available today take a week to process DNA samples, whereas the Ion PGM produces results in a couple of hours, says Rothberg.The technology is also what makes the system cheap and scalable, he says, comparing it to a personal computer. "As you democratize it, it gets cheaper for everybody," he says.

The technology will not compete with established platforms such as HiSeq, made by sequencing giant Illumina of San Diego, California, which costs about $700,000 and can read 250 billion bases in a single run. Instead, with its initial capacity of 10 million–20 million bases per run, the Ion PGM will probably be most useful for microbiology labs and clinical researchers, who work with moderately-sized samples.

The HiSeq's capacity is overkill for a lot of labs, says Rob Knight, a microbial ecologist at the University of Colorado, Boulder. Each run costs thousands of dollars and processes thousands of samples. Being able to run small amounts of pilot data cheaply — and to ask a question and get an initial answer within hours — "would transform what people can do and probably make the pace of development much greater", he says.

Realistically, the ancillary costs of running the machine will add up to more than its $49,500 price tag. For example, buyers will also need to purchase a $16,500 computer server for data analysis. Mitchel Sogin a microbiologist at the Marine Biological Laboratory in Woods Hole, Massachusetts whose lab won an April competition to receive the Ion PGM for free, says his lab will now need to acquire a special cell cytometer to use with it.

An expanding market

The Ion PGM sequencer is not the only new technology set to enter the sequencing game. Pacific Biosciences, based in Menlo Park, California, is beta-testing a much-awaited machine that it plans to launch next year. But in the current early phase of the market, platforms are likely to have differing strengths and drawbacks.

Sequencers from Pacific Biosciences, for example, won't be priced for individual labs, but the technology can read 500 base chunks of DNA at a time — a capability important in deciphering certain areas of a genome — compared with the Ion PGM's 100. However, at an investor meeting last month, Pacific Biosciences noted that its machines were so far performing at an accuracy of just 80–85%, whereas Ion Torrent says that its machines can achieve 99% accuracy.

As you democratize it, it gets cheaper for everybody. Jonathan Rothberg , Ion Torrent

Jonathan Eisen, a microbiologist at the University of California, Davis, says that his university's genome centre has ordered a machine from Pacific Biosciences. For his own lab, he has gone for a GS Junior platform from 454 Sequencing in Branford, Connecticut — a benchtop machine that relies on an older technology but costs about $100,000 and can handle long reads. If the Ion PGM is accurate enough, says Eisen, he may order it as a "possible complement to the longer reads".

Knight says that he needs to see more data from the company to be assured that the Ion PGM works as promised. "There is a desperate need in the field for instruments that reduce the cost" of sequencing, he says. "It's just that the companies that have promised that in general have not been able to deliver it yet."

Better and cheaper

Low cost may not be enough to push uptake in clinical or medical diagnostics lab, says Jay Shendure, a genomicist at the University of Washington in Seattle. The Ion PGM's price tag may help labs that want to get into sequencing, he says, but "instrument cost is only one of a long list of barriers facing translational genomics". The biggest barrier is that it is still unclear how sequencing will be clinically useful. And practical questions such as informed consent issues and agreement on technical and regulatory issues could still take time to work out, he says.

What's more, says Shendure, clinical applications in particular will need a set-up in which not only the sequencing itself is speedy, but so are all of the preparatory steps. Ion PGM's sample preparation takes about a day.

Rothberg acknowledges that there is some perfecting left to do. To that end, Life Technologies has launched the Life Grand Challenges Contest, in which it will offer a $1-million prize for each of seven challenges in life-sciences technology development. The first three challenges, announced today, invite labs or individuals to get twice as much sequence data from a single chip as the Ion PGM, cut the sample-preparation time in half or double the technology's accuracy. The remaining four challenges will be announced next year.

"We don't have all the answers," says Rothberg. "What we have is a flexible, low-cost platform that we are going to keep making better and cheaper."