Stories
Slash Boxes
Comments
Search

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

DNA For Information Processing and Data Storage

Posted by CmdrTaco on Mon Dec 20, 2004 03:14 PM
from the but-how-many-mp3s-can-it-store dept.
Data Storage Science
Haydn Fenton writes "Here is an article on using DNA for data storage and even information processing. From the article, "The DNA molecule - nature's premier data storage material - may hold the key for the information technology industry as it faces demands for more compact data processing and storage circuitry. A team led by Richard Kiehl, a professor of electrical engineering at the University of Minnesota, has used DNA's ability to assemble itself into predetermined patterns to construct a synthetic DNA scaffolding with regular, closely spaced docking sites that can direct the assembly of circuits for processing or storing data.""
+ -
story
This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.

DNA For Information Processing and Data Storage 50 Comments More /

 Full
 Abbreviated
 Hidden
More
Loading... please wait.

  • It works on so many levels (Score:5, Informative)

    by antimatt (782015) <xdivide0.gmail@ORG.NET.EDU.com> on Monday December 20 2004, @03:19PM (#11139873) Homepage
    And another thing: chemically, DNA is almost heroically unchanging. It is among the most unreactive, inert molecules in the biological world. That means data integrity, a Good Thing.
    • ... chemically, DNA is almost heroically unchanging. It is among the most unreactive, inert molecules in the biological world. That means data integrity, a Good Thing.

      Good point.

      When DNA does go bad, typically what happens is that the telomeres [nih.gov] wear out, leading to cell death.

      -kgj
      • Good point indeed but you misunderstood. DNA inside the cell IS changing all of the time although changes to its chemistry are being repaired all the time. Telomere change is something else, that happens at cell division. Higher-order structure, like folding, also changes. What the parent meant is that DNA, when taken out of the cell, is very very stable with most of its primary and secondary structure remaining intact over a long long time (see extraction from Neanderthal bones). However, the point of using DNA as a scaffold for the assembly of information is not in its stability per se. It's in its ability, per its repetitive structure with lots of nice modifiable side chains available, to direct assembly of other molecules. This is what is meant, methinks.
        • However, the point of using DNA as a scaffold for the assembly of information is not in its stability per se. It's in its ability, per its repetitive structure with lots of nice modifiable side chains available, to direct assembly of other molecules.

          Thanks for the clarification.

          -kgj
            • We've gone from "heroicly unmodifiable" to "has modifiable sidechains" with two +5 informatives. Someone's right.

              You have forgotten Slashdot Rule #6: Don't let the facts get in the way of the moderation.

    • On the otherhand the world is filled with enzymes called DNase that will willingly eat your TB DNA storage in no time.

      Siliconase however does not exist.
    • Disclaimer: IANA[Molecular Biochemist/Geneticist]

      I'm not so certain that's something to brag about in this case. DNA may be stable, but DNA replication is not always reliable and accurate. Genetic mutations are common - they are the result of random errors in the replication process. Some organisms have turned really rotten replication accuracy into an advantage (e.g. HIV, which mutates so fast that it has demonstrated an amazing ability to survive everything science has thrown at it). Other organisms
      • Well I quite certain they will not use 1 dna strand to store the data... they will use lots. See it as RAID but then with really cheap disks... and millions of them. As long there is no heavy selection on the replication (there is a bit of selection: AT bonds only have 2 H-bonds while CG has 3), it should be very accurate.

        btw HIV is RNA (retro) virus

      • Re:It works on so many levels (Score:4, Interesting)

        by jhoger (519683) on Monday December 20 2004, @04:00PM (#11140388) Homepage
        The solution to such problems is redundancy. One efficient form is padding with extra bits to add error correcting codes.

        Science fiction may have an answer too. I believe the Slavers [Niven, Known Space series] engineered giant food animals (with intelligence just because the Slavers were really mean) that had specially engineered DNA so that they would not be impacted by radiation. As you say, mutation is necessary in evolving systems, but if one were engineering a system, you'd want to take that out of the equation.

    • Viruses (Score:5, Funny)

      by goombah99 (560566) on Monday December 20 2004, @03:41PM (#11140170)
      So the equivalent of a SQL, insert field command, will be a retro-virus? Will my database be down...with a cold?
    • It is among the most unreactive, inert molecules in the biological world

      The backbone of DNA is very stable, but the bases definitely do not fit your description (indeed, I smell troll). Bases can be oxidized (drink your anti-oxidants), thymines can be fused (UV light causes this), etc. The reason DNA retains such high fidelity over such a long time is because there are enzymes in the nucleus that are specifically designed to repair these unwanted changes (for example, 8-oxoguanine is repaired to guanine

  • 640K DNA molecules... (Score:3, Funny)

    by GillBates0 (664202) on Monday December 20 2004, @03:20PM (#11139887) Homepage Journal
    ought to be good enough for anybody.

    -GillBates0.

  • So how long... (Score:5, Interesting)

    by JossiRossi (840900) on Monday December 20 2004, @03:20PM (#11139889) Homepage
    How long until Religious Nuts start claiming to see hidden messages encoded in our DNA telling us to love Jesus?
    Or
    How long until spies pass messages along in the form of biological matter by sneezing into a tissue?
    Or
    How long until we can buy books in readable vials full of liquid?

    The possibilites are endless and cool but of course it will probably just be used to sell us Coca Cola... so much wasted potential.

    • Re:So how long... (Score:5, Funny)

      by savagedome (742194) on Monday December 20 2004, @03:27PM (#11139972)
      How long until Religious Nuts start claiming to see hidden messages encoded in our DNA telling us to love Jesus?

      Each DNA strand is a number. Like the Hebrew A, Alef is 1. B, Bet is 2. You understand? But look at this. The strands are inter-related. Like take the Hebrew word for father, 'Ab' - Alef Bet... 1, 2 equals 3. Alright? Hebrew word for mother, 'em' - Alef Mem... 1, 40 equals 41. Sum of 3 and 41... 44. Alright? Now, Hebrew word for child, alright, mother... father... child, 'Yeled' - that's 10, 30, and 4... 44

      • Re:So how long... (Score:4, Insightful)

        by leroybrown (136516) on Monday December 20 2004, @04:17PM (#11140576) Homepage
        For the love of god, who in the hell thought this was anything but funny? Look people, the universe does not give a rat's ass that we count to ten. That is, the fact that we have ten digits in our numbering system is COMPLETELY arbitrary. So is the fact that we have 26 letters in our alphabet. I guarantee you all that if we had four fingers on each hand like the Simpsons characters, we'd be counting in octal. Anyone that manages to apply any sort of relation between letters and numbers to come up with some sort of code which bears any sort of relevance is deluding themselves.

        Think about it this way... the Bible was originally written in Hebrew. The Hebrew alphabet has 22 letters. Do you really think that the Christian God (should he exist) thought far enough ahead to include people in the bible who numerological name conversions would apply in English? No, that's absolutely absurd.

        This is why I swerve to hit these wacky numerologists out there. They're so blissfully ignorant of the arbitrariness of our numbering system and number of letters in our English alphabet that they try to apply some grandiose scheme to letters in order to convince people that they're some kind of mystic.

        • Do you really think that the Christian God (should he exist) thought far enough ahead to include people in the bible who numerological name conversions would apply in English?

          Although I absolutely agree with you in regards to these crackpot code schemes, your reasoning here is flawed. The Christian God, as taught in the Bible, is omniscient (all-knowing) as well as sovereign (in control). He is outside of time and knows exactly what the future holds. Therefore, He certainly could think "far enough ahead

  • I remember hearing about this originally nearly 10 years ago now. I remember bringing it up in a discussion on Usenet, engendering many "It will never happen" trolls...still seems a few years off though from consumer product?

    • Re:The things people doubt (Score:2, Informative)

      by Anonymous Coward
      This has (essentially) nothing to do with older ideas of storing data in DNA sequences. The DNA is being used here as a scaffold to lay down a particularly dense array of unformatted storage material.

  • SCO and DNA (Score:4, Funny)

    by lordsilence (682367) * on Monday December 20 2004, @03:22PM (#11139911) Homepage
    Just wait till SCO find infringing code in YOUR DNA..

  • Slow Posting (Score:5, Funny)

    by ClownsScareMe (840001) on Monday December 20 2004, @03:24PM (#11139931)
    Could people be *gasp* reading the article?

  • Nasty unforeseen consequence (Score:5, Funny)

    by borroff (267566) on Monday December 20 2004, @03:24PM (#11139940) Journal
    Unfortunately, when the tune "Jingle Bells" is coded in DNA for storage, it turns out to be a version of the flu...
  • DNA is just a biologic/chemical process of storing info. The smallest bit of information you could reach has already been hypothesized to be an electron...polarize it one way and make it positive (one) and the opposite (zero). Last time I checked electrons are smaller than DNA. But could we go smaller? Quarks? Neutrinos? Photons?...as the smallest components of information?
    • You can store information on anything, as long as you keep the temperature low enough or the energy high enough. The minimum energy per bit you need is kT/6, where k is Boltzmann's constant and T is the absolute temperature of the medium where your information is stored. Sure, you can store a bit of information as the polarization of a free electron. But unless the temperature is low enough, your information will be quickly randomized by thermal noise. Or else you can have your electron in an atom, inside a

  • DNA plus Information Processing equals... (Score:3, Funny)

    by amstrad (60839) on Monday December 20 2004, @03:32PM (#11140050)
    ...Eddie, your super friendly shipboard computer.

  • Performance Limits on Chemical Computation (Score:5, Informative)

    by tjic (530860) on Monday December 20 2004, @03:34PM (#11140083) Homepage
    For those interested in such things, a friend wrote his PhD thesis on defining the limits of achievable DNA computation:

    Performance Limits on Chemical Computation [dyndns.org].

    • Wow, excellent link.

      The Slashdot-relevant portion of the abstract:

      I compare the maximum possible performance of proposed DNA computers from the literature with current commodity electronic computers, and conclude that diffusion-driven, DNA hybridization based computers cannot exceed the performance of current electronic computers by more than a factor of 40,000, and probably by much less.

      This doesn't speak to using DNA as a construction scaffolding, which I am not skeptical of, but DNA computing has never

  • I would prefer to use RNA (Score:4, Funny)

    by dfn5 (524972) on Monday December 20 2004, @03:43PM (#11140190) Journal
    Sure, go ahead and use DNA, if you want your data to mutate.
    I will use RNA (Raided Nucleic Acid) instead.

  • Carl Sagan pointed out in Dragons of Eden (Score:5, Interesting)

    by museumpeace (735109) on Monday December 20 2004, @03:45PM (#11140218) Journal
    [late 70s] that DNA was the only persistent data storage media nature had until we apes invented languages that we could symbolically preserve. All that has essentially progressed, and what has been changing rapidly with advances in biotech, is the speed of data access into DNA. 5 yeas ago, the best guess [and the big money of govt and industry] was that it would take us 10 years to transcribe the human genome...and now thats already done. We are getting faster even faster than we expected. [that technological acceleration could be partly attributed to the open exchange of techniques and discovered sequences that the consortium of biochemists had agreed upon at the outset of the project...kind of like developing products in open source]
    When that data access speeds up another 8 or 10 orders of magnitude and is both R and W,[and not much sooner!] we can talk about DNA as if it were magnetic media and seriously talk about its applications...Makes you wonder if the lessons of open source are going to have to be rediscoverd as we further exploit what software engineering has to teach us about handling DNA.

  • DNA computers (Score:4, Interesting)

    by wronski (821189) on Monday December 20 2004, @03:46PM (#11140228)
    There has been some discussion about using DNA as a massively parallel computer. Suppose you encode data in a DNA sequence (input), then somehow act on it (running a program), and then read the resulting altered DNA. You have a computer, albeit somewhat slow and not terribly practical. Now imagine you start with not one but *billions* of different DNA sequences.You "run" the program over all these inputs simultaneously, and obtain billions of possible outputs. You can then use some chemical tag that binds itself to the 'correct' answer. You now have a massively parallel computer with negligible power consumption in a test tube.

    This sort of DNA computer could be useful for a number of problems that involve a lot of trial and error, such as protein folding. In a paper some years ago some scientist managed to solve a traveling salesman problem using one such computer. They generated different strands corresponding to each city, and let them mix in a tube randomly to produce different candidate 'paths'. Then, they used some chemical selector (the tricky part) to eliminate the strands corresponding to invalid paths. Left in the tube were all valid paths, which could then be easily replicated using PCR.

    I couldn't find the original paper, but a pretty good explanation can be found here [howstuffworks.com]

  • I can just hear it now... (Score:3, Funny)

    by Bhasin_N (838449) on Monday December 20 2004, @03:47PM (#11140254)
    "In Breaking news, a minor short circuit has caused a freak mutation at Genetic Information Inc. causing all the chips to asexually reproduce and take over the coffe machines"
  • Art imitating life: In ST:TNG a Klingon was found to be stealing secrets by reading information off some chips. The raw data was encoded into inert DNA chains and eventually injected into a person. In effect anyone could become a roaming hard drive and not even know it.
  • First of all, it is super-twisted. That means that you've massive latency in accessing it. Secondly, it is linear, which means that finding the Nth "word" is going to be HORRIBLY slow. Thirdly, the existing duplication system is "good enough" only on a large scale. Finally, the protein generator has no exception handler, so if the raw materials aren't available, there is no means of handling the exception.

    Having said all that, if you've massive archives where you're less interested in immediate access as

  • I just finished reading Richard Dawkins' The Selfish Gene, and was thinking about some stuff like this. Note that I'm a computer engineer and don't know much about this stuff:

    Evolution has worked for billions of years. I'd say it's worked out quite a few of its bugs. So why don't we copy it when doing computing? I think the book stated (VERY generally, I assume), that there are 4 bits that get used to build with DNA - A, T, C, and G... obviously this has advantages, so why are we using binary computers

    • The reason we use binary computers is because computers are based on transistors operating in saturation mode, and long-term storage is based on alignment of magnetic materials or optically reflective pits. I suppose we could have computers whose basic unit of data had more than two levels, but it would be a lot less reliable than our current technology until some significant problems could be solved (such as getting these non-saturation-mode transistors to work well at very, very high speeds).

      On the othe
  • The technology could help computers identify objects in images with something approaching the speed of the human eye and brain

    Man, always thought my computer had no trouble winning against my brain in terms of speed.

  • By the way, if you want some good info on home hacking of DNA, proteins, etc. check out DNAhack.com [dnahack.com].
    • Just because the information is in my body somewhere doesn't mean I can access it during an exam. Otherwise I'd have gotten perfect scores on every exam. My point is that you just need to prohibit DNA reader devices, then all the DNA cheat sheets in the world won't help you any more than all the Spanish I've learned helped me during the exams...
    • Your post shows a complete lack of understanding of DNA completely. Please don't post about this again.
    • wow this makes me shiver when I read it.

      1. it's spelled protein.
      2. It can only be a, c, g, t in dna
      3. those aren't proteins they are bases
      4. Proteins are coded by dna, they do not comprise dna in any way.
      5. Dna doesn't get spliced during cell replication. It gets copied.
      6. therefpre isn't a word.

          • > .the peptides (yeah yeah, not protiens)

            No they are not peptides either. They are simply bases. You are right though NULL doesn't work. The fact that they come in pairs doesn't limit this in any way. You always get 2 strands of dna they pair with each other. So long as you only look at one strand there are four possibilities in each position.
            • > any molecular biologist want to correct

              Well a biochemist of sorts would be better. As a genetecist I'll respond. A pairs with T and C pairs with G. Dna is a double helix and one is an exact inverse copy of the other. Where one has an A the other has a T and vice versa. Since generally only one strand is read you have base 4 at each site.

All trademarks and copyrights on this page are owned by their respective owners. Comments are owned by the Poster. The Rest © 1997-2009 Geeknet, Inc.