My students in study hall today asked for help with science homework. The question was, “Is the probability of two persons having the same DNA small enough for a jury to convict a suspect? Explain your answer.”
We got the bit about identical twins, but they wanted to know what the chances were of two apparently unrelated people randomly having the same DNA. We found an FBI website that stated the chances of this happening were 1 in 260 billion. Of course the next question was, how long would it take for 260 billion people to have lived on Earth? Not at the same time, obviously, but how much time would have to pass for 260 billion to have lived and died on Earth? They want to figure out how long it would be before they would randomly have an identical twin pop up. I’m not exactly sure how to find this on the Internet or in research, so I was hoping someone would help me answer this question for my students. Thanks.
Well, the 260 billion figure does not make the DNA identical. It just makes the DNA that they check identical.
DNA identification schemes use polymorphism analysis – these are sequence changes that occur frequently throughout the genome. For identification, they pick a bunch of different sites across the genome and look at the specific changes. The frequency 1:260 billion comes from two people being identical at these specific places.
There are thousands of polymorphisms known in the genome, with more discovered every day. Large scale polymorphism discovery projects are being carried out across the world. These are what account for differences between individuals. The chance of two humans being identical at all loci is basically nil (except in the case of identical twins and even then I’m sure that if one looks hard enough, one can find differences). If they are checking twenty or a hundred different polymorphisms, and they just happen to line up with a rate of 1:260 billion, it probably won’t be reflected at all in how a person looks and certainly doesn’t mean they are identical twins.
Cecil answered a related question, but I don’t think it matters. If there are 6 billion people on Earth and the probability that two have identical DNA is 1 in 260 billion, the expected number of people who have DNA identical to yours is 6*1/260 = .023. And that’s ignoring what edwino said, so the actual expected number is even lower. So yes, it should be good enough to convict.
Yes. While juries have probably never been presented with DNA evidence of sufficient resolution where the probability or two persons having the same DNA is relevant, juries have been presented with DNA evidence and returned a conviction.
This probably says more about juries than the quality of DNA evidence, though. Current methods are not good enough to isolate to a single individual, though it is a positive tool for excluding suspects. I wouldn’t convict on the basis of DNA alone… and if I was hearing a case where I’d expect DNA evidence to play a role (e.g. rape), I’d want to hear a damn good reason why none was presented.
ALso, keep in mind, that with DNA tests, not only do they only test specific sequences, they need to have good source material. I have read of some cases where a less than stellar blood sample could lead to the fact that the blood could have come from any genetic member of a family, and from there other evidence was needed to indicate guilt, but the DNA did certainly limit the pool of potential supects.
Another problem, which we are running into big time here in Houston is contamination. The Harris County crime lab was in utter disarray, and therefore all DNA samples passed through there are now suspect. This goes beyond the mundane mislabeling and sample switching. Most of these polymorphism detections depend on exponential amplification of the DNA in a procedure called PCR (polymerase chain reaction). In theory, one can get a good PCR product (enough to see by gel electrophoresis, so at least several nanograms) from a single strand of DNA. I routinely do a 2[sup]35[/sup] amplification maximum (35 cycles of a chain reaction). So even a little bit of lab sloppiness can cause PCR contamination – for instance a lab tech sloughing some skin cells into a PCR tube just by touching it without gloves on, or worse yet contamination of some of the ingredients that go into the reaction mix.
IIRC, this is the line that Barry Sheck (sp?) took when arguing against the DNA in the OJ trial. He called Kary Mullis, the inventor of PCR, who has been overly critical of the protocol due to its high risk of contamination. In practice (at least in my practice ) contamination is easy to detect and usually easy to get rid of. Run your negative and positive controls and if you see bands in the negative lanes, just getting all new reagents and clean your bench and wear gloves usually does the trick.
) contamination is easy to detect and usually easy to get rid of. Run your negative and positive controls and if you see bands in the negative lanes, just getting all new reagents and clean your bench and wear gloves usually does the trick.