Commercial, USB-Powered DNA Sequencer Coming This Year

Zothecula writes "Oxford Nanopore has been developing a disruptive nanopore-based technology for sequencing DNA, RNA, proteins, and other long-chain molecules since its birth in 2005. The company has just announced that within the next 6-9 months it will bring to market a fast, portable, and disposable protein sequencer that will democratize sequencing by eliminating large capital costs associated with equipment required to enter the field."

Yes, goodie

[Theaetetus]

So when can we expect to see one in every police cruiser, insurance office and personnel department?

More importantly, we can expect to see one in every doctor's office and hospital, allowing inexpensive personalized medicine.

Re:Yes, goodie

[Professr3]

It wouldn't be just one. They aren't reusable, so it's going to cost $900 per sequencing operation - apparently, you have to throw away the whole device afterwards.

Re:High error rate

[Anonymous Coward]

I don't think you understand how DNA sequencing works.

Re:High error rate

[the gnat]

All of the "next-generation" sequencing technology has a relatively high error rate compared to traditional Sanger sequencing (used on the original human genome sequences, and still the gold standard for truly novel genomes). The massive redundancy typically compensates for this, although some technology is clearly pretty marginal no matter how much data you have. If my memory is accurate, the Human Genome Project was collecting somewhere between 6x and 10x redundant data; projects using the newer tech shoot for more like 30x.

What I don't get is what this device is intended to be useful for if it's only able to sequence 150 million b.p. before wearing out. The article mentions that this is smaller than some human chromosomes, but unless they factored the necessary redundancy into that figure, it's not going to go very far. It'll be enough to sequence most bacterial genomes, and probably enough to sequence human cellular RNA transcripts, or something else targeted, but I just can't see it being useful for whole-genome analysis of the sort that tries to answer deep questions like "am I likely to get Alzheimers in twenty years?"