sequence an entire human genome at that speed and cost, Smith says. One run should generate 150 megabytes of genomic data, based on the blog reviews, which means it would take about 720 runs to get a decent human genome, at a cost about $360,000 and 720 hours of work, Smith says. That really means the Ion Torrent tool isn’t able to crank through as much data in the same amount of time as other machines, making it lower in “throughput,” in its current form, he says.
“So that’s not competitive with the other platforms for whole genome sequences, but it is practical for many other applications. And it can go into labs more easily,” Smith says.
Still, the tool from Ion Torrent is likely to be useful in many ways, Smith says. It might be used by researchers who want a quick, easy genomic readout on other organisms like bacteria or viruses, or certain regions of DNA that are of interest in a specific human tissue, Smith says. Because of the low cost, speed, and simplicity for the user, it could lend itself over time to diagnostic applications, Smith says. And importantly, the low purchase cost of $50,000, compared to $500,000 for other machines, could enable Ion Torrent to “democratize” sequencing by selling its instrument to a broader pool of biologists than just the small group of people who run hard-core sequencing centers. That could unleash the creative juices of many smart scientists who don’t really have easy access to sequencers but could do cool things if they did, Smith says. The machine also is likely to increase its throughput over time, as semiconductor speed and power increases.
The news from Ion Torrent is really just the latest step in the world of DNA sequencing, which has been on a breakneck pace to make sequencing better, faster, and cheaper. Illumina announced in January that it is now possible on its machines to sequence entire human genomes for as little as $10,000. Mountain View, CA-based Complete Genomics says it can do the same job for $5,000, largely because it has a different model in which it doesn’t sell machines, but rather it asks researchers send their samples to a centralized company lab. We’ve also written about other companies that are developing machines that aspire to push the leading edge of speed and cost, including U.K.-based Oxford Nanopore Technologies and Providence, RI-based NABsys.
It all sounds great. But like anything new, Ion Torrent will have to prove that it can do what it says it can do, Smith says.
“There were a lot of people there who were saying they are going to take a serious look at this,” Smith says. “But you also had the world’s best sequence instrument operators there. They’ve built large centers, and they’ve heard everything before from every company. They’re going to ask, ‘Where’s the data?'”
