Sunday, 23 August 2009

Musings on DNA.

It all looks the same, really.
(Picture swabbed from here, where you can see just how expensive DNA analysis really is.)

Some commenters on this post were much more up to date with modern DNA technology than me. I'm a microbiologist who uses some DNA methods, not a molecular biologist who is expert in them all.

So it was especially interesting to hear from The Paragnostic in the comments there, who gave a description of the actual test used on those samples. It's far simpler than I thought. All it involves is the use of a specific enzyme to break the strand at specific points. Where those points are will differ between individuals but will not differ much between close relatives, if at all. The number of bits left, and the sizes of the bits, gives the 'profile'. It's not really very specific at all, given the number of people our government want on that database.

The potential for your sample to match with Ronnie Kray is rather higher than it would be if they were testing actual gene sequences. Cross-matches will happen. Instead of a database of criminals, where a DNA sample might throw up two or three matches, we'll have a database of millions of people which will throw up hundreds, maybe thousands of matches from a sample. Not much help if you're trying to narrow your search. They're building a haystack around a small pile of needles. Are they doing that because a) they are all astoundingly stupid and nobody has realised the problem, or b) because catching criminals is not, and never was, the purpose of that database?

This test will not distinguish between a human and a chimp. Since the lab will assume that the sample is human DNA and probably do no supplementary tests (such as checking to see if the sample contains human antigens), then the test might not spot the difference between a human and a banana.

So, coat your cheeks with banana and give your name as Mr. Fyffe. Then wait for that dawn raid on the greengrocer.

Seriously, this does lead to a rather more worrying scenario, which comes from a comment by Anonymous on the OH post.

My lady has just come back from a night in the small town of Cirencester which has recently seen some trouble.
Apparently there is a large police presence and they are going around swabbing people..........

Well, my first thought was 'just another database-filling exercise' but then I thought about it some more. Those police taking swabs are not scientists. They take the swab, demand ID, stick the ID to the swab and send it off. They are not trained to think about potential problems in sample collection.

Now, suppose they swab you. So what? You've done nothing wrong. You aren't concerned about being mistaken for a kiwi fruit when the next crime happens. You aren't worried about matching Gary Glitter's profile. Nah, not worried at all. Your ID is tagged to your DNA and that's that. Because you're carrying your real ID.

A criminal might not be. A criminal might have fake ID. Very likely, they bought their fake ID from an identity-thief who has picked up thousands of real peoples' ID details from bins and phishing attacks. In a random DNA-collection exercise, the criminal is the one with nothing to fear because their DNA sample ends up tagged to someone else's ID. They don't need to run or hide or protest at the sample collection at all - in fact they'll be first in the queue. The DNA belongs to the criminal. The ID tagged to it does not.

It could be yours, and you won't know until one early morning when the battering ram smashes your door in.

'Oh', you might say, 'it won't matter, as soon as they test my DNA they'll see it doesn't match'.

Will they test it? Why would they? It's already on the database so why test it again? The evidence is already there and DNA is infallible so why re-test?

The criminal has nothing to fear from the DNA database. The innocent have plenty to fear from it.

Whether you have anything to hide or not.

One more thing. For those who think handling and extracting DNA is really hard - I first extracted human DNA from a cheek swab in 1979, in a biochemistry practical class, first year. Every student got it right first time. It's very, very easy to do. That's without modern automated machinery. So don't believe the 'it's too hard for a criminal to do' crap. Given a swab of your cheek cells, I can get the DNA out of that in my kitchen. Certainly enough to leave a good sample at a crime scene. Really, that sort of brute-force framing is a doddle.


Anonymous said...

More & more worrying. Love your DNA piccie BTW. I keep promising myself a visit to the City of Arts & Sciences in Valencia where they have a great deal about the Human Genome Project (I think it's called that) & a huge model strand of DNA which you can see through the window as you walk past on the pavement.

Rogerborg said...

Oh, do give over. Sure, the DNA database is just fantasy jizz-material for the Stasi, but you give the criminal class way too much credit.

I'd recommend that you read some plod and magistrates' blogs: they seem uniformly of the opinion that most lags are from the shallow end of the gene pool, and are barely able to remember their real identifies, let alone fake one up.

The main danger of false positives isn't enemy action, but good old fashioned bureaucratic incompetence. Never mind getting your record out of the database - good luck even getting a correct record into it.

Anonymous said...

A while back there was a case where a DNA fragment amplification technique "pioneered" by UK forensic biologists was used and there was a fair old hoo-ha about both the validity of the technique and the statistical framework supporting the inferred result.

From an scientifically literate (molecular biochemistry) outsider's point of view it certainly looked like one of those instances where the outcome of the experiment is determined by the answer being sought - life's to short to chase every red herring - but it all went quiet about the "revolutionary British technique pioneered by Government Scientists" Russian genetics sprung instantly to mind - wonder why?

Tom said...

Rogerborg has a point but - well, there's been some talk - and I for one - don't know the truth of it.

If you get swabbed and databased and you were totally innocent as in part of a sweep and the perp is collared - later - if you go to get a job as a school caretaker or a visiting children's author - does the fact that your DNA is on record flag on one of those checks? And... is this deemed sufficient grounds for denying the job?

Also - extrapolating from Rogerborg's understandable take on the British summary justice system - a bureaucratic cock up - not like we've had any at all eh, let alone ones involving databases - I keep getting flashbacks to Terry Gilliam's Brazil.

Now, if we could get it re-released.....

caesars wife said...

you can hear the labour stasi thinking weve cracked it this time .

Chances of possible error 1 in 2 million

ergo: find some at crime scene you must have been there

reinforce this ultimate tell tale evidence in court cases and in media , and you lay the foundations for a DNA database that in the future would enable you to pick out criminal perpetrators on a few specks left at the scene .

Police appear infalable , state knows where you are , state can enact thought crime in the media as actual crime is pretty much unthinkable .

except no one thought to ask if either the database or the evidence could be corrupted for political purposes or even criminal ones .

Just in case you think the post is a bit one sided , DNA evidence has so far correctly identified a number of old crimes where criminal left a good sample , it can help in elimination of suspects as well.

but courtcases where they are only going on tiny samples and no other corrobative evidence are erm a bit lite .

so DNA evidence is fine in making cases with other evidence . DNA database too much near to stasi thinking for my liking .

sconzey said...

Not to mention the fact that cheek swabs are a poor way of taking DNA samples.

Can you think of any reason people on a night out might have someone elses DNA in their mouths? ;)

Anonymous said...

Agree that the main danger is the bureaucratic incompetence.
But when you say this is giving the criminal class way too much credit, you're kidding - right?
Plod and magistrates opinions will be based on those who are always having their collar felt. The clever ones avoid this exam by not getting caught.

Anonymous said...

Instead of a database of criminals, where a DNA sample might throw up two or three matches, we'll have a database of millions of people which will throw up hundreds, maybe thousands of matches from a sample. Not much help if you're trying to narrow your search. They're building a haystack around a small pile of needles.

The core material for those who practice eugenics. Targeted, engineered, pandemic flu viruses for instance.

A.B. Gordon said...

DNA "magic" during the Omagh bomb enquiry someone in Nottigham came up with a "perfect match" for the bombers DNA. Door kicked in MI5s "finest" pistol whipping...O crap it is a 14 yr old schoolboy with no connection to Tyrone.

formertory said...

The DNA database's reliability and accuracy is of course completely immune from effects of falsely-generated profiles, as well as from incorrect handling by Plod....

Anonymous said...

I smell a wizard wheeze in the offing. Wouldn't it be fun if if Mr. defence lawyer stood up in court and proclaimed:
Actually, your honour, the samples matching those found at the scene of the crime exactly match those of Bob the chimpanze presently detained at XYZ chimp rescue centre as well as the accused's.

Leg-iron said...


There is no 'criminal class'. It's the only occupation that is entirely classless.

Did Bernie Madoff have a sloping forehead and drool?

Did he have access to limited funds?

He wasn't al that smart. The smart ones cut the scam in profit, change identity and vanish to a country where nobody cares about your past as long as the palm grease flows.

Seriously. There are slope-headed monkey-boys out there committing crimes and they get caught all the time.

There are people who could join MENSA committing crimes too.

They don't get caught.

Forget the 'criminal class' idea. There isn't one.

Rogerborg said...

True, but the monkey-boys vastly outnumber the Madoffs, and I know which ones are a threat to me and mine. Drive with me through any Welfare Wasteland of your choice and we'll play a game of 'can you spot the criminal class'.

As you say, they already get caught all the time, because:

1) They are morons.
2) The Criminals' Justice System operates a revolving door policy that keeps letting them out to get caught again and again.
3) As a consequence, they (rightly) have no fear of it. Kindness does not restrain the wicked.

Either way, the DNA database is largely irrelevant to crime, because it won't help with the Madoffs, and it's not needed for the mongs, and just sucks money out of enforcement and imprisonment budgets.

Ratings and Recommendations by outbrain


Related Posts with Thumbnails