Site Meter
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
Gene Expression Blog
The Intel of Sequencing
Credit: Forbes

Credit: Forbes

Matt Herper’s piece in Forbes, Flatley’s Law: How One Company Is Creating Medicine’s Genetic Revolution, is a must read. Here’s the conclusion:

It’s hard to disagree with him. The cost of sequencing a human being’s DNA is less than one-hundred-thousandth of what it was when Flatley started running Illumina 14 years ago. Illumina is hoping to lower the price further. Ronaghi, the CTO, says the market has been disrupted every time the cost of sequencing has dropped by five to ten times. He foresees DNA sequencers that might cost $10,000, as compared with $250,000 for Illumina’s midline models, opening up whole new markets–and cures. Says Flatley: “The road maps that we have are pretty breathtaking as far as where the technology can move in the next three to five years.”

As narrated in the piece competiters have tried to take Illumina down a notch since 2008, and failed. So you don’t want to bet against them. That being said its dominance is having an effect on the famous chart of “cost per genome” decline. With the stabilization of the monopoly prices haven’t really dropped much since 2012. It might take a full frontal assault by a credible competiter like Oxford Nanopore to induce more than incremental change on the margins.

• Category: Science • Tags: Illumina 
Hide 11 Comments

11 Comments to "The Intel of sequencing"

  1. D. K. says:

    Is this referring to something other than our current ability to get our own DNA decoded– through, say, 23andme.com- for less than a hundred bucks each? Because, if that is what it is referring to, it strikes me that it is simply conflating Moore’s Law– which applies to the processing power per dollar of the processors themselves– with the market prices on offer for an end product that merely utilizes those processors. When 23andme.com, for one, chose to lower its price to $99 (and even less, for additional samples on the same order ), that was a marketing decision, based upon hoped for economies of scale– not a matter of a particular breakthrough in available processing power per dollar, at the technological level.

  2. 1) the chart is for sequencing. not snp chips

    2) not really emphasized, but my understanding is that personalized genomics in the domain of cancer is going to be a big area of growth/interest

  3. D. K. says:

    Okay, thanks for clarifying that for me, Razib!

  4. D. K. says:

    P.S. I also should have mentioned, in my original comment, economies of scope– as I believe that 23andme.com’s real motivation, in lowering its price to under $100, was its hope of thereby amassing a large enough customer database to be able to market that database to other companies.

  5. The graph looks to me like the price is still falling at pretty much the same pace as Moore’s law. Which could be monopoly pricing – that Illumina doesn’t think the industry can expand fast enough to drop prices any more than that.

  6. #5, the period where it went way beyond moore’s law may also be an aberration. though if oxford nanopore’s claims are to be believed then we’ll see a second crash in the next few years.

    #4, yes, been well known that that’s 23andMe’s aim in selling the tests at below market.

  7. Biologist says:

    Re: the chart. The underlying source is here: http://www.genome.gov/sequencingcosts/

    The cost per genome metric for the last 2 years shows relatively flat costs during this time period:

    Apr-12 $5,901
    Jul-12 $5,985
    Oct-12 $6,618
    Jan-13 $5,671
    Apr-13 $5,826
    Jul-13 $5,550
    Oct-13 $5,096
    Jan-14 $4,008
    Apr-14 $4,920

  8. gwern says:Website

    Biologist: I downloaded that data and played around with a bunch of curve-fitting (piecewise linear, exponential decay, splines, and a few others; I can post the R code if anyone is interested) but I’m having a hard time coming up with any plausible model using the data from the last ~5 years in which it can cost $5k to sequence a genome in April 2014 and August 2014 Illumina can boast about a $1k genome. Do you know what’s going on there? Is this some sort of accounting issue where the NIH data-series is a ‘total’ cost and Illumina’s playing number games to get $1k?

  9. Illumina says “$1000 genome, inclusive of instrument depreciation, sequencing consumables, DNA extraction, library preparation, and estimated labor for a typical high-throughput genomics laboratory.” I don’t see any games there.

  10. Biologist says:

    The $1000 figure may be based on the HiSeq X platform, which is currently sold in instrument 10-packs.

    http://www.illumina.com/systems/hiseq-x-sequencing-system.ilmn

  11. Current Commenter says:

Comments are closed.