Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Encryption Security

DNA Encryption 102

Posted by Hemos from the better-living-thru-you dept.
brn writes "There is this BBC article that talks about hiding information inside the DNA. Very interesting and well worth a read, apparently information hidden this way is virtually undetectable. " It's the espionage agencies that first got interested, but the notion of "watermarking" is another that is brought up-and you thought PIII ids were bad.
This discussion has been archived. No new comments can be posted.

DNA Encryption More

DNA Encryption

Comments Filter:

  • Re:DNA Storage Capacity (Score:1)

    by Anonymous Coward
    It would all depend upon the length of the DNA strand. You have four bases, so each base could (in theory) represent two bits of data. So I would imagine that you could fit a good bit of information onto a long enough strand-- but I'm not sure how long a strand you could embed into the DNA of an actual human being without causing problems-- i.e., cells interpreting the embedded information as the bases for a protein, etc. But I'm sure a little research will turn up answers.

    Incidentally, it might also be possible to embed information through other means:

    When creating a DNA fingerprint, enzymes are used to "slice" a strand of DNA whenever a certain sequence occurs. The result is a whole load of smaller DNA stands, each with different lengths-- the exact number of strands and their lengths is unique to each person/animal/whatever. A small electric current is then used to pull the miniature stands across a special material. The smaller strands travel more quickly across this material than the large ones, causing a unique fingerprint that looks something like:

    || ||| | | | || ||

    If an organization (say, maybe, the NSA) were to sequence a strand that were to be cut at certain points by the right enzymes, they could (in theory) create a strand of DNA whose fingerprint represents a bar code that holds data.

    Any other ideas for data hiding in DNA?

  • I like the BBC (Score:1)

    by Anonymous Coward
    Another interesting article by the BBC. The Real Audio file is good too if you check it out.

    I think European media in general is so much better because they are sponsored by the state. The European states have checks and balances to make sure the media is not a propaganda tool. Our American media does not have the same safeguards to prevent commercial interests from twisting the news.

  • Re:Prety scary stuff.. (Score:1)

    by Anonymous Coward
    Espionage (at least the popular meaning of the word) tends to be more oriented towards obtaining information that you shouldn't have. So an agent infiltrating an organization and obtaining their five-year plans would be espionage -- but how he sends it back to his home agency is SIGINT (signals intelligence). Unless, of course, he just walks... :)

    In some sense, this brings new meaning to the old joke about BTP (bipedal transfer protocol)...

    Tampering with human DNA, though, isn't as bad as it sounds. The demonstration in the article doesn't involve any tampering with DNA in an organism -- a letter that you write and send (yes, some people still do :) is going to have a little of your DNA on it; so mixing a message sequence in and putting the cypherdna on the letter isn't an issue.

    As for humans: it's not like that one Far Side where a kid is sitting on an airliner, and his armrest contains a "Wings Stay On/Wings Fall Off" switch. The human body would interpret a little random junk in the occasional DNA strand as just that -- random junk. It won't kill anyone, make them grow a third arm, or even turn them into Steve Gutenberg (halleluia!) Keep in mind: (1) They're only affecting one out of every 30 million (something like that) DNA strands, and (2) they're not going to give you a pill that transforms every DNA strand in your body.

    I'd imagine that, if a human were to be used as a carrier, they'd doctor up their DNA appropriately and then inject a small amount just under the skin (like a tuberculosis test). The solution would stay for a few days before dissipating throughout the body, so there's no effect on the "packet" (if you will), and it's easy for the intended reciever to extract and decode the message (whereas an enemy would have no idea where to begin, and couldn't do anything nasty to find out before the solution dissipates). So, even though it sounds a bit scary, it's not so bad. (How much worse is it than high-strength encryption?)

    BTW - I'm not sure about capacity. Something like 90% of our raw DNA is thrown away and not used to make proteins anyway: if the data were placed in those portions, we could be talking about data capacity in the tens of thousands of base pairs. A base pair translates to two bits, (assuming you have a scheme to tell which backbone is the one you want), so if we use 10,000 base pairs, we can encode 20,000 bits (or 2.5 KB) of data.

  • Exactly what we need (Score:1)

    by Anonymous Coward
    Great, so even though we still really have no idea how DNA works (we do know it's not quite as simple as "codons 3 billion through 3 billion and four determine if you're blonde"), we're going to go using it for steganography. DNA is not like jpegs, people. We can't be sure there's such a thing as a "least significant amino acid". What if it turns out that writing "This person owned by Gene-Corp" means "cancer of the armpit" in cell-speak? Then again, perhaps we'll have some sort of wacky scenario in which a Johnny Genetic courier suddenly manifests Akira-like psionic powers because he's carrying Yakuza bank account numbers in his DNA. Heh.
  • It sounds just like chaffing & winnowing [mit.edu], except using DNA instead of packets.
  • What amount of data could you store in DNA with that method? I guess this may be the first true molecular storage device. However it is hopelessly redundant since it has many DNA strings all with the same sequence stored on them. But considering the amount of data needed to encode a human being, I imagine you could possibly put audio/video streams in a sequence. Would take a while reading it back though, forget about streaming it.
  • Gee. Why don't cattle ranchers stop branding their cattle and just keep the animals' information in a big centralized database?

    People don't crash or erase their own DNA. Not to mention the space that would be required to store that volume of information in a database.
  • They don't need to sequence 30 million strands of DNA. PCR amplifies out a target DNA fragment using oligonucleotide probes which are specific to regions of DNA. After the PCR has been performed, the purified PCR products may be sequenced quite easily, in a single step.
  • hmmmmm....so now we can watermark our leach-neuron claculators with all sorts of info :)


    Who am I?
    Why am here?
    Where is the chocolate?
  • I agree that that is what the article said, but what is to prevent them from actually inserting that DNA back into a cell? From what I gather from listening to news reports on sheep cloning and such, you can't insert DNA into adult cells but you can insert it into embryonic cells. And, of course, there has been fetal tissue research as well. Would it be possible for someone to introduce the modified DNA strand into fetal tissue and then place that tissue into an adult human?


    Who am I?
    Why am here?
    Where is the chocolate?
  • That's exactly what I was thinking....one of the worst science fiction movies of all time is about to come true! ;-)

    What's next?

  • It could be placed in the interons. With sufficient care it would do no biological harm there.

    Risks? Sure, plenty of them. But since when have intelligence agencies been concerned about risk, especially when it's someone else at risk.

    More specifically, the risk is that the interon sequence may end up active if an unlucky data stream is coded there.

    I'm not too worried about that since there's not much (if any) advantage to embedding the coded sequences into an organism's genetics vs. simply placing the DNA on something. For example, simply spraying the coded DNA on a courrier (possably an unwitting courrier) and collecting a sample of that courrier's dead skin later.

    Example, sequence your message. Suspend the encoded DNA in a liquid carrier in an atomizer. Go to the airport and spray a suitable tourist. Tourist goes and has a nice vacation. Agents at the destination collect DNA from bedsheets at the hotel. Tourist is none the wiser, and not in any danger unless enemy agency finds out they're carrying a DNA code sequence.

  • The genetic material you already carry is sufficient to uniquely identify you. There's no need to add a watermark.

  • ... yeah, right, and Gulf War Syndrome was sold by US Biowar companies to Saddam Hussein back in the '80s while he was at war with Iraq, and used on US troops in Desert Storm, and now the USAF is spraying US cities with a genetically modified virus to impart US populations with resistance to the Syndrome so we don't catch it from the infected troops returning home. . .

    http://strangehaze.freeservers.com/index.html


    "The number of suckers born each minute doubles every 18 months."
    -jafac's law
  • I think there have been a number of Sci Fi stories based on this premise, that the ancient aliens that visited prehistoric earth encoded messages in our DNA, meant for us to decode millenia later when we had advanced far enough -
    One of the best variations on this (IMHO) was The Andromeda Strain, where aliens had sent out millions of microscopic space "capsules", with a genetically engineered germ living on them, designed to live off the material of the "capsule", and radiation in space. It would be cheaper to mass produce billioins of these capsules with the germ, and send them out into space, than to attempt to broadcast radio waves into space to contact other civilizations. Since these capsules would normally burn-up on entry to an atmosphere, a non-spacefaring race would not be able to receive this "signal", but in the story, Earth sent up satellites to scoop micrometiorites out of orbital space for research. The craft that intercepted the capsule crashed in Piedmont, Arizona, and the germ mutated, and destroyed the entire population. The USAF then took the capsule to a secret underground lab where they picked it apart and discoverd the secret. (but not before some tense moments when the thing mutated again, and started eating all the rubber gaskets, and leaked into the lab, and the automatic nuclear self-destruct was triggered - just as the scientists figured out that this thing thrived on gamma radiation.


    "The number of suckers born each minute doubles every 18 months."
    -jafac's law
  • . . . and this idea comes from the X-Files. . .



    "The number of suckers born each minute doubles every 18 months."
    -jafac's law
  • The article is talking about DNA spread around on surfaces on the outside of cells... but the idea in many SF plots is to incorporate messages into inheritable DNA - I vaguely remember a storyline where introns ("junk" segments) in Human DNA are found to contain coded messages from the beginning of evolutionary history, put there by aliens etc. etc.

    Or get this, for a way to pass a message undetected:
    Incorporate a secret message into the DNA of a highly contagious (but non-fatal) flu virus, infect yourself. Spread the disease around by copughing in the subway, etc.. Disease spreads around the world, and is picked up by destination lab, sequenced, decoded and message is delivered. It's a tradeoff; slow delivery time vs. a untraceable delivery path (i.e. no radio or message handoff).

  • For once there is a technology that they cannot patent.

    No, the technique itself can't be patented. However, bizarre though it sounds (at least, it does to me), individual DNA sequences can be (and routinely are) patented.

  • The BBC is of course state sponsored, by that nasty television tax. The biggest advantage of this has to do with the fact that the UK government forces the BBC to provide programming which the market would not have otherwise done. How big this advantage is, occupies debate. If the market would not have paid for the programming, the audience to see the programming probably doesn't exist either. On the other hand, it does allow some good shows to be produced and perhaps one of them may be a sleeper that suddenly has a tremendous impact on people.

    However, privitization of the BBC is always talked about. These days, people in the UK, like here in the US, have access to satellite/cable television. Since there exists a good market for quality programming, there will be a channel with that programming. Competition is hurting the BBC, especially with regards to news. The news.bbc.co.uk web site is outstanding, and it is a result of severe competition in the news industry...huge investments were made to make it possible.

    The thing is, the British, in general, have a higher demand for good quality news than Americans do. I need not remind people that the magazine "The Economist" is outstanding, and is not state sponsored.

    Quality local news happens in the US, and it is the result of markets. A report is available which shows the percentage of households watching news in metropolitan areas. Columbus, Ohio, my current home, is rather low on the list, despite the size of its metro area population, consequently, the news isn't all that great. On the other hand, the other "C" cities--Cleveland and Cincinnati, are both in the top 10, and the news programming is more diverse and higher quality. Good broadcast jouralists strive to be in the higher rated cities, and it is no coincidence that Cincinnati has been a point on the way for some to become national anchors (including Jerry Springer, which is another story.)

    My point is, Americans seem not to demand higher quality news (and it is not like it isn't available...for instance, the Macneil News Hour on PBS is very good, but underwatched. NPR is excellent, and is on par with BBC radio.)

    I am not sure about the checks and balances statement. The BBC is more curtailed in coverning British government than American news organizations...because of tradition and law.

  • > If you have cable TV, see if you can get BBC News 24.

    Why bother getting up to switch the TV on? Use their their " BBC News On-Air [bbc.co.uk]" Internet service. (You can even get Radio 4's wonderful "Today" news show in Real Audio format [bbc.co.uk]!)

    Regards, Ralph.

  • I'm as naive as most others when it comes to biology, but it seems as though it would be unlikely to have tagged DNA in a living being. The reason steganography works in, say, image files is because there are unused bits in there. I know of no such unused space in DNA; once you change it, you have a fundamentally different string. What does that add up to in a living being? Cancer?

    An expert clarification would be appreciated.
  • what is to prevent them from actually inserting that DNA back into a cell?

    I am not a biologist, but I imagine that the standard cancer-destruction mechanisms would kill that cell.

    What appeals to me as a cryptogeek is the notion of finding or creating two separate genomes, which when recombined (or mapped and XOR'ed) revealed a message.

    You could do quite a good short story on this, depending upon the means of recombination. Human beings probably would not make very good one-time-pads, but I'm sure that something could be thrashed out.

    "like an infinite number of monkeys, there exist two creatures whose DNA, then XORed together (after using the appropriate decoding scheme to generate a bitstring)" will generate {the complete works of shakespeare, the meaning of life, the text of (insert your favourite holy book), god's last message to his creation}..."

    lotta potential there.
    - alec

  • That's not actually true. Only retroviruses actually insert stuff into the host's geonome. Something like influenza doesn't do this, it just takes over the cellular machinery.

    Well, recombination can occur without any specific mechanism -- just at a low frequency. There's nothing stopping it. Horizontal transfer is emerging as one of the more important causes of molecular evolution, in fact
  • It beats having a barcode tattoed on your arse ;)
  • If you really want to categorise people like this, why not just have a big database with everyone's DNA fingerprint in it, with an attribute on each object that says what category it's in?
  • I think the human gut is pretty good at handling random foreign DNA - after all, that's exactly what it does with the animal and/or vegetable DNA that you ingest every day... Viruses usually come protected by a protein sheath that enables them to penetrate cell walls and (if they are retroviruses) get into the nucleus. My cell biology's a bit rusty but I think this is right :)

  • Re:Scary (Score:1)

    by Cato ( 8296 )
    I don't claim to be expert, but something like 90% of all human DNA is 'junk' (now called something else I think), i.e. not known to code any proteins. This so-called junk DNA may in fact have some subtle purpose, but as long as the 'watermark' was coded into junk DNA it should be fine.
  • For some reason this made me think of chaffing and winnowing, in which a series of messages are sent, each with a message authentication code (cryptographic checksum) that marks the message as authentic (i.e. you can check that it was signed with the sender's key). Any messages with fake MACs are noise that is filtered out by the receiver. It was described on the Net, and you can find a description under rsa.com somewhere.

    The proposed scheme is really steganography, but if you encoded a number of messages into DNA, you would just need to attach MACs to the real messages and fake MACs to the fake messages, adding significant security.

    I guess the marker that lets you pick out the signal from the mass of human DNA is rather like computing the MAC and finding that it matches the included MAC of a message.

    Of course, you could just encrypt the signal messages instead - chaffing and winnowing is really intended for when you want to 'encrypt' without using any encryption, so to speak (only authentication is used to achieve confidentiality here.) However chaffing and winnowing helps by making the actual signal messages very hard to detect.

  • Hey everyone, this is not encryption. Why? The message itself is never obfuscated or transformed in any way, instead it is merely hidden among a lot of other DNA, on a very small dot that wouldn't be noticeable under brief scrutiny. The only similarity between this and encryption is the existence of a secret (the base-pair sequence of the DNA strand "caps") between two parties.

    There is a term for this type of message-hiding, but it's eluding me (someone around here knows it, I'm sure...). Other popular examples include:

    • Putting a message into a .GIF or .JPEG image by slightly altering the colors of some of the pixels (where the system of alteration is known to both the sending and receiving parties).
    • Putting a message into a very small dot of ink (like the period at the end of a sentence) on an otherwise benign piece of correspondence. The receiver can see the message under a microscope, if he knows where to find the dot.

    Again, this is different from encryption because the important message is hidden in an otherwise benign and useless message. So the hard part of breaking the system is knowing whether a message is hiding another message. Once you know that, breaking the scheme is pretty trivial - there's no computational complexity behind the system.

  • With that type of encryption, Langley, VA would
    probably look like the bar from the "Star Wars"
    where "to be decrypted" people would hang out and
    have some innocent, harmless fun, like killing each other. That would establish one to one
    correspondence between killing the messenger and
    cancelling the message.
  • Reminds me of a short story by Kelvin (John Walker) "We'll Return, After This Message" [fourmilab.to]
    Not great fiction, but it is interesting to read it ten years later in light of this story and the SETI at Home news.
  • Unless of course, someone wants to mark you as theirs.

    (All _my_ slaves have the same genetic watermark. ;-) )

    --

  • The problem with this method is an ability to synthesize long stretches of DNA. For now only 100 nucleotides (it is enough for 20-30 latin letters) can be made in one set. To write a one page report would cost enormous amount of equipment time and money.
    This is amazing that Nature published this paper.
    The idea is on a surface. For years one letter code is used to define aminoacids and stop signals and people have fun putting their names in DNA for a long time.
  • 1) First, you can do it too. check out the DNA-o-gram generator:

    [clearlight.com]

    Encode your own secret messages in DNA code. Now all you need is a synthesis machine to create your encoded ladder, and someone to give it to whose got some good biology knowhow and a gene sequencer machine. (I'm sure there are plenty of people in column A, but not column B)
  • From the limited information in the BBC article, it seems pretty unlikely that this type of encryption will be widely used. The expense and effort that would go into:
    a) modification/creation of desired DNA sequences
    b) DNA analysis
    would make this technique prohibitive. And, in terms of security, how many people would you trust in these labs anyway? (A one-man operation would probably have trouble handling many samples ...)

    And, as mentioned elsewhere, this technology is nowhere near the point where they can insert coded messages into your own genome! They could, however, put the information into self-replicating circles of DNA (plasmids) that can be propagated in microorganisms ... this would give you other ways to send your encrypted message (in food, tiny vials, a handshake even!). Of course, this would still be expensive and time-consuming ...

    (Before I began my developing career I was a molecular biologist, and stories like this always catch my eye ... Now I will have to read the actual Nature paper to see the details!)

    YS
  • The correct term would be steganography.
  • I don't know squat about cryptography, but I find this cool. I had actually thought of writing a science fiction story along these lines a while back but well, blah, blah, . . . I didn't.

    The cool part about it is no one expects a message. If you send someone a CD, papers, e-mail, anything than it suspect. But if it's in DNA, inside someone, they pop in, leave a drop of blood or a skin scrape, voila message delivered.

    Or how about using someone (something? . . . dogs, cats) as an information repository. They walk around with all this data encoded in them and who would know? *They* might not even know.

    One thing that's pretty cool is you can layer information in DNA sequences by making a one base pair frameshift. There are viruses that use this technique to pack as much genetic information into there small head capsules as they can.

    It looks like this:
    ACG ACG ACG ACG
    CGA CGA CGA CGA
    GAC GAC GAC GAC

    Of course, that's if your using 3 base pair codons, it might be easier to encode using straight base pairs. Which would allow for more storage? You might have to use natural codons though, in order for them to exist/be replicated in your snazzy new BioDrive (TM).

    . . . Act now!! Store all your pr0n, ROMs and MP3s in the privacy of you own DNA, never leave home without them again!!
  • The method you suggest would probably be far easier and practical to use than straight base pair encoding. Although, you would be able to store less info per strand, it would be relatively quick and easy to read it.

    This is pretty far fetched, but how about storing information in the three-dimensional aspects of proteins. We don't have the means of calculating the shape of a protein from its amino acid sequence yet, but if we did ...

    How about storing information in the number/position of cell receptors, then you could do some kind of immunoassay to search for certain limited messages.

    Your idea is definitely more feasible. I had actually thought of writing a science fiction short story in which espionage was committed in much the same way as described in the article, but what held me up was where would you store the DNA? You have encoded your message, the people at the other end have their gels ready, you put the message-DNA in the carrier ...where? It would have to be somewhere it wouldn't get digested but you could retrieve it for decryption, any ideas?
  • The only use of watermarking would be to mark groups of individuals, but even this seems to me to be rather useless since I would think it would take longer to check the DNA than to do a background check.
  • How exactly do they know that you have a TV to charge you the tax. Is there some form to fill out or does it just work more on the honor system.
  • Who, it seems to be possible that this method, used during the progress of cloneing, allows to mark a human during his whole life without any chance to remove it...

    Normaly I'm luckey if people find another way of data-storage. But this expectation realy scares me.
  • I sure hope none of the goverment branches will use this for espionage... somehow the idea of tempering with human DNA seems a bit spooky to me

    The human DNA is removed a person and mixed with the coded DNA just to present background nooise. The DNA isn't biologically active and in fact would get chewed up by polymerases within biological organisms.
  • There is a very simple way to insert DNA into (all) cells of a mature human beeing: a virus.

    That's not actually true. Only retroviruses actually insert stuff into the host's geonome. Something like influenza doesn't do this, it just takes over the cellular machinery.

    Even with retroviruses it's a bit difficult to get the virus to inject itself into the geonome. For example, the lambda phage lyses the host E. coli cell most of the time and integrates itself rarely. By fiddling around with the lab conditions, you can encouragely integration but its still difficult. The preferred vectors are stuff like an F element(E. Coli) or artificial chromosomes(S. Cerevisiae) due to size considerations.

    For humans, you would need to limit the size of the message due to the size limitations of the vector. Plus, with current technology you would have to wait a generation or two to get a person who had the genetic code in every cell.

  • Could be used to identify you as Artificial (clone, replicant what have you) or as belonging to MS!!! Trademarks on oil eating microbes etc.

    Interesting, maybe not as scary as some people think but a little disturbing...

  • These vans are not capable of detecting a single tv set in your room. This is technically impossible (your tv is only recieving programs, not sending. Their equipment is not that sophisticated to scan the low emission level of your tv). All they are doing is to test the quality of terrestrial broadcasting.

    In Germany we are having the same system of a monthly tv fee, collected by the GEZ (Gebuehreneinzugszentrale). You are expected to pay that fee, and they sometimes send people to households who don't, trying to find out whether you own a tv by asking trick questions or something like that...
  • It looks like the end markers tell where the message is and the message itself is not encrypted so much as just remapped to a different alphabet. The technique of hiding messages is called steganography (not encryption), and this instance would have to be the best example of "security through obscurity" I've ever seen! ;-)
  • Regarding the practicalities of this technique: I imagine it could be rather simple to extract the message if someone developed a convenient lab kit. They might have the result in an hour.

    But creating the message - how could that be done quickly? Doesn't it take some time to create a specific DNA sequence? I can't imagine we can do that quickly outside a lab.

    If this holds, the best use I can imagine is in "fingerprinting" items. Imagine them used like "taggants" in explosives. First you create a a whole library of different sequences (a precomputed alphabet) and then assign them to specific items or specific lots of items you want to track. One hundred dollar bills, for instance. The detection lab kit would then become a standard part of the counterfeit detector's toolkit.

    Anyway, my point really is that this new thing is really useful in the signature scenario than where you want to send an arbitrary message.

  • OK, in crypto language what we have here is a symmetric, secret key system where the key is the marker sequence at the ends of the DNA chain (I'm assuming the coding of the message into four bases is trivial). To put it into a more commonplace form, imagine taking your plaintext, encoding it into, say, the lower bit of 3 seconds of an audio CD recording, putting it into the middle of an audio CD and then sending it to the recipient. The key is knowing which 3 seconds to decrypt. Accordingly, you need a secret key which tells you: when you see a, say, 0x6a6c7ff45c054 sequence, start decoding. The advantage of using DNA is that your "CD" is very very long, but you can compensate for this in a variety of ways.

    So, while this whole thing sounds cool, cryptographically this is nothing new and has no particular advantages (all the problems of dealing with secret symmetric keys, etc.). No wonder the NSA was not interested.

    Kaa
  • You and a lot of other posters forget that there is absolutely no need to watermark people -- everybody is already "watermarked".

    Consider this: to watermark a person you need to access his DNA and insert a watermark, then access it again to read it. Well, if during the first time you just read his DNA (which is unique) and the second time read it again and compare to a database -- voila! exactly the same result without all the unpleasantness of invasive DNA surgery. Why do you think reasonable people get worried when the police decides to take DNA samples from everybody arrested? (I don't know if the law/regulation to that effect passed in New York, but was definitely considered).

    Kaa
  • Comment removed based on user account deletion
  • Comment removed based on user account deletion
  • Comment removed based on user account deletion
  • Hey, Asimov's from June 1999 has a story called "Written in Blood" by Chris Lawson that covers this idea. Highly recommended.
  • This seems like something out of a Neal Stephenson book, pretty neat. I have to wonder about the stuff they can do that they *are not telling us about* You can bet that there are some geek scientists in an undergroud bunker doing working on something that would make your head spin.

  • And encoding "Microsoft" in a persons genes will make them a mindles Micro$oft drone.
  • I believe the added difficulty here is that the DNA must be sequenced. Sequencing 30 billion strands of DNA takes a very long time, and this would be required before applying computational search techniques (distributed or otherwise). But, if the marker is known before sequencing then the DNA that is interesting can be filtered out in the lab.
  • I agree with your piece regarding the hype except the last bit. Cryptography has much to do with information theory and DNA biosequence analysis uses statistical techniques from machine learning to try to identify features in the DNA (e.g. exon-finding or promoter recognition).

    DNA statistical models could be used in more clever ways than in the BBC article to encode messages that looked like other DNA. This is not difficult and could be done today.

    Even better would be to exploit the biological machinery by creating a message in a synthetic gene that is expressed in the presence of some regulatory element, perhaps a synthetic small molecule.

  • That's just my point!
  • This news was on slashdot a while ago (a week or two maybe). Not this specific article, but talking about the same paper they're talking about.
  • There is actually quite a lot of "unused space" in the human genome. The actual percentage of DNA triplets that code for amino acids in proteins is relatively small (don't recall quite how small) because lots of big strings of DNA actually get excised during RNA processing. Other big stretches of DNA are used as receptors for proteins that determine when the gene is expressed and things like that- other times, long sequences are used for structural changes in the DNA. What's more, there are also long long strings of pure "junk" DNA that's the equivalent of ABABABABABABABABABAB etc., with no apparent purpose.

    So actually, there's quite a lot of unused space. The kicker is, there's not a limit on how big the genome can be. As long as the cellular machinery can handle it, there's no reason that you couldn't just insert sequences of however long you want into the DNA in the bits that get excised out or in stretches of junk DNA. For a change in DNA to affect the organism, you need a lot of other supporting DNA to ensure that the gene gets turned on- otherwise, it's just like a big string of garbage characters in a book- you just skip over it, it doesn't make reading the book any different.

    Now, what's interesting is how compact viral DNA is. Viral DNA is so efficient (since so much information has to be packed into a tiny space) both "sides" of the DNA can be read to produce separate proteins. Normally, only one strand, the coding strand of DNA can be read- the non coding strand actually functions as kind of a check for errors. Viruses dont' care that much about accuracy, I guess... Here's what it looks like:

    --Gene B
    3'TGACTAGTACGTCATGAAGTCAGAGGGTC5'
    5'ACTGATCATGCAGTACTTCAGTCTCCCAG3'
    Gene A-->

    Fascinating, huh?
  • that last bit didn't come out quite right- that'll teach me to preview!

    --Gene B
    3'TGACTAGTACGTCATGAAGTCAGAGGGTC5'
    5'ACTGATCATGCAGTACTTCAGTCTCCCAG3'
    Gene A-->

  • can't seem to learn the difference between submit and preview, can I?

    I'm assuming my inadequate HTML knowledge is screwing up my arrows, so I can't draw the positions correctly.
    That Gene B thing should be over to the right, with an arrow pointing to the left.

  • Re:DNA Storage Capacity (Score:3)

    by hey! ( 33014 ) on Thursday June 10, 1999 @06:02AM (#1857671) Homepage Journal
    According to the human genome project, they will be establishing markers every 100K (1x10^5) base pairs, and will need 30K (3x10^4) of these markers.

    That works out to 3 billion (3x10^9) base pairs.

    Each base pair is made up of one or two pairs of units; if we distinguish sense from antisense strands in the DNA, this means we can encode four values (or two bits) of information per base pair, working out to 6x10^9 bits, or a little more than 5.5 GB (note that since only one side of a strand is used in life, this means you can spell "human" in the genetic sense in less than 2.5 Gbits).

    As somebody else pointed out, this scheme is not encryption, but a method of information smuggling. That is, instead of using it to send secure messages over an insecure channel, the technique is supposed to hide the fact that a message even exists.

    Although virtually undetectable by casual inspection, it's hard to think of a practical application of this technology because of the effort and equipment needed to extract the message. Sending detailed plans for sabotage to your terrorist cell this way seems impractical, unless your terrorists are biologists with lots of time on their hands.
  • It's just a nit that I'd like to pick,
    but without knowing the PCR primer "password",
    it's unlikely that you'd be able to amplify
    enough DNA to read its sequence.

    So, no: The DNA is not exactly sitting there
    "in the clear" and able to be read. Not only do you need to know where it is, you also need to know a little bit about what it is.


    But, yeah, technically you're right: The message isn't encrypted. Encoded, certainly, but not encrypted. Maybe if you define "encrypt" as "hide" but I digress.
  • Point taken. Nonetheless I think people around here need to be hit on the head with a "Not Everything Can Be Reduced To A Facile Computer Analogy" stick.

    How about a coding sequence for an autoendopeptidase that post-translationally cuts itself up into fragments whose noncovalent interactions reveal the plaintext?

    Or maybe I should just stop sniffing glue.
  • This article is so content-free, it hardly
    bears comment except for the hysteria it seems
    to be inducing.

    First of all, the technology described here --
    synthesis of DNA oligomers and use of the polymerase chain reaction (PCR) to detect
    them -- is old hat. Every molecular biology lab in the world has been using these techniques
    routinely for over 15 years.

    Second of all, coding DNA into something
    biologically "meaningful" is orders of
    magnitude more difficult than spelling out a message.

    Finally, you all already can be uniquely identified by your DNA. Nobody needs to implant an ID in you, Big Brother can just take a drop of blood from you and store your DNA fingerprint in a database if he thinks you're worth looking after.

    In closing, I would urge everyone not to look at biology (or even bioinformatics) through a hacker's eyes -- it doesn't work; the two fields don't map to each other well.

    Oh, PS: If you want to see something genuinely cool being done with DNA by a hacker, check out Adleman's (the "A" in RSA) tackling the travelling salesman problem with DNA oligomers. (Sorry I don't have a cite handy.)
  • I think it's called steganography.
    Stick this word in your favourite search engine
    and see what comes up.

    Someone just sent me a message-hiding-in-a-gif
    program yesterday. Apparently it can
    encode a whole website inside a GIF picture.

    We will need this in Australia because of the new
    Internet censorship law.

    Cheers,
    Alan Kennington
  • Haven't seen this addressed before, but considering we haven't completely mapped the human genome supposedly, isn't there a sort of danger that our hidden data might cause some problems whether inheritable or accidentally causing a disease or something? Anybody have any idea?
  • the sci-fi movie i was thinking of was Terry Gilliam's Orwell-esque "Brazil" the gripping black cult classic. home of the infamous Ministry of Information, who still manage to blunder through to a bent and bizarre beurocratic end despite a shitload of blundering enroute. it all begins with a typo...

    this DNA watermark story has all those hallmarks. to the point it relies on alphabet conversion to a piece of DNA spelling out " the secret message in the middle, plus short marker sequences at each end" then these synthetically engineered strands are just "slipped into a normal piece of human DNA".

    gee i hope their data entry staff have high accuracy levels! no room for error...and what? do you meet these encryption chicks in bars, they lure you back to their place to privately encode you? or boys for us girls?

    makes james bond and bond girls look postively small town. but probably far more likely.
  • The term you are describing is ste ganography [nightflight.com].

    As to whether this article describes encryption, it could be interpreted either way. I've looked up various definitions of encryption, and it looks like you could end up in a semantic mess of codes and symbols if you think about it too hard (after all, written language is just another code). It comes down to whether the intentions of the encoder count. If they do, it's not encryption, since encoding the message in DNA appears to be merely an expedient. If they don't, it is encryption, since this transformation of letters into DNA bases has the unintentional effect of making the message unreadable to a stranger (who somehow happens to bypass the steganography).

  • When I read the article correctly, it doesn't say anything about transferring the DNA inside a living being, it is sent on the back of a postage stamp, glued on the dot in a letter, etc. They just use some random human DNA to make the information-carrying DNA 'invisible'. Later, the information carrying DNA is filtered out by using the special markers at beginning and end.
    So, no cancer here.
  • Hmm, I don't see much use in DNA as a message carrier. You can't artificially synthesize more than a few hundred base pairs at most, and even Nature's best polymerases can't do more than a few million without making a mistake. Nor is it possible to use PCR to amplify a message of more than few thousand base pairs. Reading the message, by means of a sequencer, will also be a slow and painstaking process.
  • Every type of media has the possiblity of getting damaged, that's why we have CRCs and error correction algorithms.
  • In last month's Asimov's Science Fiction magazine there was a story called Written In Bloodabout someone who develops this kind of a technology and uses it to inscribe passages of the Koran into the blood of believers. Since this would be passed down from parent to child it could be used to mark Muslims in a way that was impossible before. It raises some interesting ideas about uses for writing in the genetic code of humans. I recommend the story and the magazine in general.
  • There is a very simple way to insert DNA into (all) cells of a mature human beeing: a virus.

    viruses do nothing else - they paste their payload DNA into the infected cell.

    You only have to put the modified/watermarked DNA into a harmless, but widespread virus (influenza, Bio-Weapon engineers have dozens of carriers at their hands) and there you go...

    The only question is: why should any1 do such a thing ? your genetic code is already unique...

    --
    Jor
  • I was thinking the same thing. In fact, I was wondering just how big the key space actually is. Since all one really has to do is figure out which markers are being used, the key space isn't the 30 billion they were talking about, but probably a more reasonable number which it might even be possible to break with distributed processing ala www.distributed.net [distributed.net]


    As far as the watermarking goes, that is not new technology at all. I remember reading the same stuff many years ago when researchers first started cloning mice. The scary question to me is: how do we know people, plants, animals, etc., aren't already being watermarked?


  • No, you're missing the point of watermarking. It's not just reading a person's unique ID and matching them, it's the ability to brand people/animals/whatever into categrories.

    Everybody considered a "hacker", "gay", "lesbian", "christian", or whatever could be marked with the appropriate permanent brand.

    What if a variant of herpes (or similar genetic disease) was created which would alter the recipient's DNA to carry this watermark?
  • Methinks you know something of the business. BTW, sequencing is still pretty expensive too, running about $500 a gel which gets you order 50K usable bases read 100Kbits... that's if you're good at it and do it a lot.

    Why bother when I can quietly slip as much data as I want into the noise of a compressed image file and hide it in the massive flow of pornography on the usenet alt.binaries. conferences?
  • So, how do we know that someone or something hasn't done this already? Like, maybe we're just a carrier for a message, a carbon-based animated post-it that some other critters will read when they arrive, like "nice real-estate in this quadrant of the galaxy". Or maybe we are the message, the expression of some higher order of process... like Alan Watts said, "the universe is peopleing." I kind of like that one. Life forms as artistic creation of 4.5 billion year old massively parallel molecular supercomputer.

    Too much. I have to go lie down now.
  • Found out this thread by checking my referred logs (I have the DNA-o-gram generator at http://dna2z.com/DNA-o-gram [dna2z.com] (thanks Infiniter X).

    1) DNA encryption is really no different from any other nonobvious code-based encryption. You must still transmit all the information to decode the message in a secure manner. The method referred to in the BBC article operates in the same manner as the DNA-o-gram generator (try it out), except that the code is generated randomly. If someone were to intercept the DNA-to-text code and the information necessary to isolate the coding DNA from the "noise" DNAs, they could crack the code.

    2) This has nothing to do with human DNA. There is DNA in everything from the fungus between your toes to houseplants. The DNA used for encryption would probably be chemically synthesized.

    3) A good sequencing run can read 1000 bases of DNA. Because three bases correspond to one english letter, you could get ~300 letters at a time.

    4) You would be better off adding fake gene sequences to a gene database and giving someone the keys to decrypt it. That saves all the hassle of synthesizing and sequencing the DNA.

    A description of Adelman's solution to the travelling salesman problem can be found at http://dna2z.com/dnacpu [dna2z.com]. (whuppy)

  • I sure hope none of the goverment branches will use this for espionage... somehow the idea of tempering with human DNA seems a bit spooky to me.
    And in the wrong hands it could become a very dangerous (+efficient) method of info transfering.

    BTW - it doesnt say how much KB you can "stuff" into one person's DNA sequence...ne idea ne1?
  • Or 'This is a previous Microsoft product buyer' translates to 'protoplasmic ameaboid'.

    Come to think of it, perhaps Bill's figured this one out already...


  • We are basically walking computers using vast amounts of data to move/think/eat/sleep - With this new method, instead of people having to go to school and waste their precious lives away learning absolutely nothing, they could, instead, have information injected into their DNA, making them evolve from a complete moron, to an absolute, omnipotent being.

    The only problem with this is that the government(s) will not allow this to happen, considering they like to keep knowledge imprisoned away from people - the goverment hates it when people get access to huge amounts of knowledge. That would explain why all our education systems suck - hardcore. Death to the United States, I say.

Slashdot Top Deals

To be is to program.

Close