DNA Fingerprinting

Professor Sir Alec Jeffreys, from the University of Leicester, discovered the process that we no call "DNA fingerprinting". Chris Smith heard how it works...

Chris - Thank you for joining us. What's involved in the process of genetic fingerprinting?

Alec - The technology has evolved quite a lot over 21 years since the first genetic fingerprint. What we do nowadays is we start with a genetic sample. This could be a little bit of blood, saliva or whatever. Firstly, we have to extract the DNA from that and then amplify up the DNA using a wonderful technology called the Polymerase Chain Reaction (PCR). We amplify lots of short pieces of DNA which we know vary a lot from one person to another. That variation is in terms of the length of the DNA. So the next stage is to separate those fragments of amplified DNA by size, measure the size of the fragments and then store the information on a database. Nowadays, most of this can be done by an automated process.

Chris - How much do we vary genetically from one person to the next?

Alec - We're all absolutely unique genetically unless you have an identical twin. In this case, you will be the same genetically. In terms of differences between any other two people, we're talking well into the many many millions of differences. In DNA fingerprinting, we don't look at the whole lot. All you need to do is to pick out ten (for the UK) highly variable sections of DNA.

Chris - How much DNA do you need to actually do a DNA fingerprint? Presumably if someone were to go through a dustbin and find a letter with a stamp on it that I have licked, they'll find DNA. Would it be sufficient in order to retrieve my genetic profile?

Alec - Yes, most certainly. We quite routinely carry out analysis on DNA from a single human cell or a single DNA molecule. We showed many years ago that you can extend that to DNA fingerprinting. It's very very sensitive.

Chris - So what are the chances of getting an incorrect result and potentially arresting the wrong person?

Alec - It's not zero. The so-called random match probability, which is the chance that your DNA profile is the same as mine, is about one part in ten trillion. Which is pretty low.

Chris - Now in addition to fingerprints, are we in the position to infer what people might look like on the basis of their DNA sequence?

Alec - There's a lot of interest in this. You can determine gender from DNA, and you can determine with reasonable accuracy whether the person is a red head or not. There is a test for eye colour, but it's pretty iffy actually. There are markers out there that can give you clues on ethnic origin, but not very strongly. That's really about it. All the kind of fanciful notions you can plough into DNA from a crime scene is, at the moment, science fiction. It's science fiction, for example in the case of facial features, because we really don't understand the genetic basis of how genes and their variation control the length of the nose or how much your ears stick out. I think there's also another issue. The current DNA systems use DNA variation which is actually pretty irrelevant to you as an individual. I think this stops there being any problems with genetic privacy. If the police start looking at very important genetic characters, such as ethnic origin or facial appearance, then I think that raises very important issues about genetic privacy, and the issue of whether the police should have access to that type of genetic information. I don't think they should.