Slashdot Mirror
Scientists Learn To Fabricate DNA Evidence
Hugh Pickens writes "The NY Times reports that it is possible to fabricate blood and saliva samples containing DNA from a person other than the donor, and even to construct a sample of DNA to match someone's profile without obtaining any tissue from that person — if you have access to their DNA profile in a database. This undermines the credibility of what has been considered the gold standard of proof in criminal cases. 'You can just engineer a crime scene,' said Dan Frumkin, lead author of the paper. 'Any biology undergraduate could perform this.' The scientists fabricated DNA samples in two ways. One requires a real, if tiny, DNA sample, perhaps from a strand of hair or a drinking cup. They amplified the tiny sample into a large quantity of DNA using a standard technique called whole genome amplification. The other technique relies on DNA profiles, stored in law enforcement databases as a series of numbers and letters corresponding to variations at 13 spots in a person's genome. The scientists cloned tiny DNA snippets representing the common variants at each spot, creating a library of such snippets. To prepare a phony DNA sample matching any profile, they just mixed the proper snippets together. Tania Simoncelli, science adviser to the American Civil Liberties Union, says the findings were worrisome. 'DNA is a lot easier to plant at a crime scene than fingerprints,' says Simoncelli. 'We're creating a criminal justice system that is increasingly relying on this technology.'"
Well, fuck.
they have to rewrite next season of CSI because of this
Company selling test to detect whether this has happened shows off a tech demonstration of why their product is necessary.
If it's for-profit but free, you're not the customer -- you're the product (e.g., the Slashdot Beta's "audience").
Ok folks, don't get yourselves in a tizzy over this.
If you read the article (yeah, I know, it's against Slashdot rules, but give a try anyway) you'll see that all this hype originates from a company that has a product to detect faked DNA evidence, that they hope to sell to forensics labs.
The simple fact is that if someone wants to plant your DNA at a crime scene, there are many possible ways for them to obtain *real* DNA to use for that purpose. They aren't going to go through the hassle of creating fake DNA...
Don't tell me to get a life. I had one once. It sucked.
Really? It's that easy? God, I'm an idiot. After I cloned the tiny snippets of the common variants, creating my library, I just sat there staring at them and thinking "What the hell do I do now?"
As a biochemistry grad student, I'd figure I'd need a month or so and could do it for less than $10,000 in materials not including a bit of a practice/training.
Materials - it costs about $0.15/base for small DNA strands, or $1/base for longer (>150 base) if you order from one of many companies. Enzymes run ~$100/enzyme good for about 50 reactions. You'd need about 5 or 6 critical enzymes. The PCR machine could be had for $500, or you could go old school with water baths and a timer. I bet I could get decent results with about $5-10,000 (not including labor, which would take a bit of time).
ï
Once you've created a library of the 'snippets' it would be almost trivial to clone up large mixed populations with the right signatures. (Trivial meaning less than a week, and a few hundred dollars).
As for price going down in the future - VERY fast. The tools to make/reshape DNA are still a bit arcane but have recently become both flexible and robust. There is an entire sector of private companies devoted to making DNA encoding & manipulation easier, faster and cheaper. Ordering 10,000-base strand now costs $1/base, but I would bet it pushes $0.10 within 5 years. Building it up from smaller (~100-bases) sequences is currently a bit of an art, but is not 'hard'. I would bet that that process will become much less arcane and therefore much more automated/programmable within the next 10 years to make that a matter of days of robot incubation rather than a month of grad-student labor.
Whole genome replication seems to mostly center around Polymerase Chain Reaction (PCR). PCR is an incredibly versitile technology. PCR machines cycle test tubes through specific temperatures, the thermal cyclers are cheap compared to a lot of lab equipment but still in the thousands of dollars. To do a PCR also requires some type of polymerase, nucleotides, some solutions, and short primer oligonucleotides. These are all items that aren't prohibitively expensive but aren't household items either.
Maybe I'm being too ACLU/tinfoil hat, but I'm getting a sinking feeling that someone eventually is going to try to slap some regulations on PCR, or at some point in the future, having access to a thermal cycler and PCR materials is going to be seen by law enforcement as a reason to be suspicious of you. And I think that would be a real crime. I could see a future where thermal cyclers come down in price even more, maybe high school kids will start tinkering around with PCR as kids from yesteryear played with chemistry sets before we decided they could be used to make bombs and should be banned.
Maybe not. Anyway, I think we should nip it in the bud if there's any hint that law enforcement starts thinking you need to have a good reason to manipulate DNA, just so they can keep their evidence unquestionably true.
1) Pass "homeland security" type law requiring people to register and submit DNA for national database.
2) Keep an eye out for political dissidents.
3) When they appear, have covert government agents commit crimes and plant "teh incontrovertible DNS evidence" of the dissident at the scene.
4) Dissident is taken out of the picture in a way that looks completely legitimate.
5) Bonus: Add extra brutality to their crimes to make the dissident (and by extension any of their ideas) less attractive to anyone else.
At the moment most (if not all) DNA profiling is done by examining STRs. STRs are specific spots in your DNA where a certain pattern of DNA is repeated a number of times. And the number of times it's repeated might be different for you from the STR at that spot from someone else.
So if you check many of these spots, you can make it extremely unlikely that someone else has all of these spots with the same number of repeats as you do. In the US they check 13 loci. And this fake DNA (the stuff they advertise as being possible to make just by looking in the database, with no original genetic material) is just a load of these loci, with the correct number of repeats in there.
The reason it isn't much of a problem is that the technological bottlenecks that made the human genome project such a money pit are close to gone now. Taking a genetic sample and fully sequencing it shouldn't be that much of a problem in the next few years (I mean you can already do it for the price of a coat. To proof against fake evidence, many other SNPs or STRs can be checked instead, as a confirmation. Keeping a list of another 13 STRs to be used as confirmation would be a good start, having the loci known but not recording the results in databases to prevent this kind of counterfeiting.
In order to 'engineer' a crime scene, to incriminate somebody by planting fake DNA, the first thing I need it a real, if tiny, DNA sample, perhaps from a strand of hair or a drinking cup. Then I use that to fake some DNA, which I place at the scene.
So can somebody tell my WTF, if I already have some legitimate DNA from the person I'm attempting to frame, I wouldn't just place that at the crime scene instead?
It helps to have the right sort of DNA. Say you want to frame someone for robbery, and you have their semen -- I guess you could argue that they are obsessive chronic masturbators and that's why their semen is all over the crime scene -- but otherwise, it would arguably raise less suspicion to find other sources of DNA.
GAAH! MY PRINTER IS ON FIRE!!! PUT IT OUT! PUT IT OUT!
The really scary one to me is the possibility of mastermind criminals framing prior criminals whose DNA is on file. Imagine a bio-hacker pedophile who framed people on those handy state lists, leading the authorities directly to the very people they suspect most in the first place.
- Michael T. Babcock (Yes, I blog)