$900 DNA sequencing USB stick announced by OxNano | Hacker News Reader

$900 DNA sequencing USB stick announced by OxNano

newscientist.com

120 pointskanzure34 comments

Here is a graph that shows the fall in human genome sequencing costs over time:

http://www.genome.gov/images/content/cost_per_genome.jpg

It's surpassing Moore's law. Looks like pretty soon we'll reach the mythical $1000 mark.

What was it that caused such a rapid change after 2007?

If I remember correctly, in 2007, the Sanger method for sequencing (which involves nucleotide-by-nucleotide, laborious, gel-based sequencing) was quickly replaced by cheaper methods. We started using laser-based sequencers, along with statistics, to determine longer sequences quicker, and using computers instead of legions of grad students.

If I had to guess, it'd be the rise of distributed computing.

Illumina acquired Solexa, and shipped the first Genome Analyzer (the ancestor of the HiSeq). Clive Brown was a central figure at Solexa during their rise and acquisition, and has been at Oxford Nanopore for the last few years (he's the one who gave the talk).

There's a lot of hype in genomics, but Brown is one of the few who has delivered, which is part of the reason why MinION is getting the benefit of the doubt. Very exciting announcement.

There are various versions of this graph floating around, most of which start in 2005 with the publication of the first next-generation (multiplexed) sequencing methods by pyrosequencing (454) and sequencing by ligation (George Church) for complete bacterial genomes. Costs came down very quickly after commercial release in 2006 of 454, ABI Solid (SbL), and Solexa (SbS-sequencing by synthesis). This graph in particular is referring to the cost of sequencing a human genome, which was first done on James Watson using these next generation technologies by 454 in late 2007 if I remember correctly.

any idea where to get that data raw?

ftp://ftp.ncbi.nlm.nih.gov/genomes/

edit: Oops, you probably meant the cost graphs. You might find related graphs from Rob Carlson.

http://www.genome.gov/pages/der/sequencing_cost.xls

It's off this page: http://www.genome.gov/sequencingcosts/

Soon they'll be paying us for our DNA! (for targeted advertisement etc.)

Isn't $900 less than $1,000?

Edit: err... DNA verus genome nevermind I see what you meant.

The keyword is human genome, not some other organism. Also the $900 product has not been released yet.

That is so cool, and the mechanics they used to read the molecules and miniaturize the equipment look incredibly clever for me, as an ignorant in chemistry. Biology is getting more and more "hackable". It seems we are seeing the same kind of evolution once happened with computers - what was once unreachable and reserved for just a couple people (big mainframes) get increasingly cheap and ubiquitous.

This blog post gives some more in depth information: http://pathogenomics.bham.ac.uk/blog/2012/02/oxford-nanopore...

They claim the accuracy is 96% and has the potential to get even better, which is even cooler given the long read length.

However, any error can be pretty upsetting and most next gen sequencers overcome this by doing many reads per base. I'm guessing that the current behemoth machines would be kept around for the time being to assemble the final, error free sequences?

current behemoth machines have errors too, often the best way to get a really polished sequence is to sequence on machines which have different type of errors, and interlace the data appropriately.

But how much are the enzymes going to cost?

I don't understand the downvotes, because it seems to be an interesting question, and the article is not very clear.

Is $900 the price for the gadget alone? Is the price for the gadget + 1 time enzyme kit? Is it the price of each repetition of the analysis (enzyme + disposables), but there is an additional initial cost (gadget)? This thing is reusable, or you must buy a new whole kit each time?

Ars had its own article on this saying that you do not need enzymes. It's forcing the molecule through a tiny pore and reading the change in an electrical signal. That said, one problem mentioned is that the error rate is quite high.

http://arstechnica.com/science/news/2012/02/nanopore-tech-br...

This article says:

"Both GridION and MinION operate using the same technology: DNA is added to a solution containing enzymes that bind to the end of each strand. When a current is applied across the solution these enzymes and DNA are drawn to hundreds of wells in a membrane at the bottom of the solution, each just 10 micrometres in diameter."

Perhaps this method use fewer enzymes than the old one.

10kbp reads would also be very useful for de novo assembly of larger genomes, particularly in repetitive regions. One company (PacBio) currently claims reads this long, but the error rate is around 15%.

Super long reads also going to be incredibly useful for looking at structural variation (deletions, inversions, translocations).

And alternative splicing.

Yeah, the 10 (100?) kb length is the huge thing here. An awful lot of sequencing bioinformatics is dealing with the implications of short read length. This problem is possibly going to go away now.

This is crazy.

The original human genome project started in 1990 and took a decade.

Now you can do it in two to six hours for around a grand?

I wonder if we'll be seeing more people public domain their genome?

EDIT: And thats just the "Compare to two decades ago" position.

The real craziness comes when you realize that this will inevitably get cheaper, and smaller, setting the stage for more advancements in this field. (I think at this point were all expecting to be able to actually turn genomes into living things before 2050.)

As to your last comment, JCVI already did that, synthesis is lacking for speed and cost as compared to sequencing today.

Meh, it's more than just synthesis. Synthesis costs are falling faster than Moores law, too; but you don't get things like epigenetic modifications or packing for free, and that tech basically doesn't exist yet.

Emphasis on "were all" as in, you and me personally.

EDIT: (Disclosure: I know very little about this subject and wish to know more.)

The biggest problem has been and will continue to be interpretation of the results, particularly in human clinical sequencing. Especially as more data becomes available with advances like this.

if you build it they will come.

I mean that the existence of a large and continuously growing data set will incentivize figuring out better ways to analyze it.

Edit: deleted my less informed comment ... some more informed discussion of this can be found on reddit:

http://www.reddit.com/r/bioinformatics/comments/pu595/oxford...

This is one time use, right?

Is it reusable? Can the protein pores be cleaned and reused?

Even if it's single-use, that cost per bp isn't terrible considering the capital costs of 454 and Illumina machines.

I bet that in 20 years, a lab-on-a-chip with integrated DNA sequencer & AI doctor will be as common & cheap as thermometers today.

These are interesting times we are living in.

Reply on news.ycombinator.com