[MUSIC] The fragments of DNA that is of interest these days for forensic DNA profiling are called Short Tandem Repeats, or STRs. Now, STRs consist of a little sequence of three, four, five, six, seven base pairs, and then that sequence is repeated multiple times. So let's take an example of the STR which is called THO1, and THO1 involves the sequence AATG. And then that AATG is then repeated and repeated and repeated. What varies from individual to individual is the number of times the little sequence is repeated. So for instance, a person might have the relatively common THO1 6,8. In fact, about 4% of people have THO1 6,8. Now, what does the 6 and the 8 mean? The 6 means that the AATG sequence is repeated in a block six times, and the 8 means that within the block, the AATG sequence is repeated eight times. Now why would someone have a six times repeat and also an eight times repeat? The reason is very simple, it's because every one of us has two parents. One of those, maybe the 6 for instance, will be inherited from one parent, and the other, the 8 in this case, would be inherited from the other parent. Now if we just look at THO1, which as I said is present in about 4% of people, that's really not highly individualised. Well, the good news is that there are many STRs that are known. So we don't just rely on a single STR, we rely on multiple STRs. And it's analysed in a way that's called multiplexing, which allows the simultaneous analysis of multiple STRs. Different authorities have decided to use different numbers of STRs. The U.S. national database uses 13. The U.K. national database uses 10. And as we saw before, the number of STRs you use improves the probability that you're talking about a single individual. If you use three STRs, then you've got odds of one in 5,000. Go up to six STRs, you're talking about one in 2 million. Nine STRs, you are talking about one in a billion. So when you get the U.S. number of 13 STRs, the odds are 1 in 100s of trillions. So we can really say that when you look at this number of STRs, you are truly individualising the sample to one person. So how do STRs look like when you do gel electrophoresis? The different STRs will come through that capillary tubing at different times, but they will come through as pairs because you have these different repeat numbers, the two different repeat numbers. So the signal you see on the output when you see these pairs, you know that one of those pairs comes from the mother and the other one in the pair comes from the father. Well, what you see on the slide here is just something that we drew with some drawing program so you can understand the principle. They really look like this. It's somewhat more complicated, but you can see how they form these pairs. But if you look closely, you will be able to pick out some which are not pairs. They are just a single spike on the chart here. That happens sometimes because it's possible that the repeat number you inherit from your mother and the repeat number you inherit from your father happen to be the same number, and that's why you get that single peak for some of these STRs. Because it's all inherited from your parents, this is how paternity and maternity testing is done. So here we have an example, and what we have done is to colour code it so you can see it clearly. So, at the top is the profile of the child, in the middle is the profile of the mother, and at the bottom is the profile of the father. And you can see that within each pair of STRs of the child, one of them comes from the mother and the other one comes from the father, and this is strictly true. Every single STR will match up in this way. If there isn't a match, then maybe that isn't the true parent. So here's an example for instance. There's the child, there's the mother and there's the alleged father, and you can see that in every case, there is a match between the child and the mother. And you can see that in some cases, for instance, the ones that are coloured in blue and red, there is a match between the child and the father. But it's not true for every case. And as you look through this profile, you can see that most of them, there is no match between the child and this person, the alleged father. Every single one must match to prove paternity. So in this case, the alleged father is not the father. We can also use this technique to prove the sibling relationship, whether someone is your biological brother or sister. But you cannot prove this relationship by direct comparison of the two siblings. The reason for this is that the two siblings may have inherited different STR repeat numbers from the two parents. There's no basic reason why they should be the same. So, in a pair belonging to the mother, one sibling might inherit one, the other sibling might inherit the same one or might inherit the different one. So to determine whether two people are siblings, you have to go back to the parents. So essentially what you're doing is proving that these two people are children of the same parents and therefore they are siblings. So let's recap. Where does your DNA come from? Half of it from your mother, half of it from your father, somewhat randomly, so brothers and sisters will have different DNA when the profiling is done. The exception, of course, is identical twins. Okay, identical twins, because they come from the same sperm and the same egg, will have the same DNA. And this causes problems. If you have a crime committed by a guy who has an identical twin, how do you tell the difference? So this is a case, a recent case from Malaysia involving identical twins, Sathis and Sabarish Raj. One of them was arrested by the police because he was carrying a large quantity of drugs. But by the time the case came to court, the police didn't know which of the two twins was the one they had originally arrested, so they could not prove which one was guilty. They were sure one of them was guilty, but which one? So, of course, we have to presume innocence. You have to presume they're both innocent, because you can't prove which one is guilty. So in these cases where you have twins, you of course have to rely on other features. There was one case I read about, where the twin was identified not from the DNA, because that couldn't be done, but because one of the twins smiled in court and it so happened that one of the twins had lost a tooth and there was a little gap in his smile. Now, this situation may change in the future as technology improves, and that is because identical twins do not have DNA that is completely identical. When the cells are replicating, when the DNA is being copied, mistakes happen. A little bit like typing errors as the DNA is written out. So, the DNA in identical twins is very, very slightly different to each between the two of them. Now, using the normal techniques of DNA profiling that are normally used by forensic scientists, that difference is not detectable, but more advanced techniques will be able to show that difference. The problem is, it's going to be more expensive. There was recently a case in Marseilles in France where one of two identical twins was accused of carrying out a series of rapes, and the DNA matched but, of course, it also matched his twin brother. So, it was not possible to get a conviction. The police in Marseilles were told that it would be able to distinguish the two twins by DNA analysis, but it would cost 30 million euros. Well of course, technology advances, and maybe soon the price will come down. [BLANK_AUDIO]
