If no two people have the same DNA, why do they say in court the the chances of someone else having this DNA are 1 in however many billion?
They don’t test every bit of the entire genome. That would take too long and be too expensive. The tests are done on smaller pieces that are good markers for the purpose.
http://www.scientific.org/tutorials/articles/riley/riley.html
That site goes on to explain PCR and RFLP testing in more detail that should help you.
Aside from what Exapno Mapcase said, identical twins do, in fact, have the same DNA. That’s why they’re identical. This would also apply to identical triplets, quadruplets, etc.
But each would have different random mutations, so probably not 100% exact, right?
Well, some random mutation in one somewhere is possible. You oculdn’t easily check, however. Likewise, there’s no absolute reason that two unrelated people couldn’t have the exact same DNA… but odds are that’s never happen and will never happen.
Do we know of any two people who are not twins that have the same DNA?
Not really. Mutations that happen in your body do not spread to the rest of the cells in your body. Any mutation that is present in all the cells of one’s body actually occured in one of the gonads of one’s parents. While a given cell from one twin may have a mutation not present in a cell from the other twin, the same thing is just as likely to happen with two cells from the same body.
No. And, although it’s numerically possible, the odds against it are truly astronomical.
Sort of a followup question - is there any person or persons who we actually know all about their DNA, or at least some version of their full genome, not worrying about mutations? Maybe a volunteer donor for the human genome project?
[Defense lawyer]So you’re saying that it is possible, correct?[/Defense lawyer]
A common defense tactic.
What’s it to you, if it’s possible? 
Personally, I think that if we’re going to prosecute people based on genetic statistics, the defense should have every right to harp on the fact that statistics are not certainties. It might be the only defense people have against co-incidence.
Of course, that requires a mathematically literate jury I realize… 
I was thinking of when the fertilized egg first splits. I was assuming errors are possibly introduced at that point.
Possible, but so far no one has actually documented that. There likely wouldn’t be much difference, possibly too little for our current techniques to pick up.
Statistics don’t tell you what IS possible, they only tell you how LIKELY that possiblity is to happen.
For instace, if you told me there is a one in a billion chance that sounds like a long shot, untill you realize there are 6.7 BILLION people in the world so that means statistically at least 6 people should have that (whatever it is)
The thing about statistics is they only tell you what is LIKELY to happen.
If I throw a coin in the air 100 times, statistically it should come up heads 50 times and tails 50 times. But if it came up heads 100 times, you would probably think the coin was fixed. But there is nothing that says it CAN’T come up 100 times head, only that it’s not likely to.
Twins don’t have identical DNA, let’s face it nothing in the world is TOTALLY similar. The thing is it is VERY expensive and VERY time consuming to test for it. So much of their DNA is identical it could a lot of time and money to find the unidentical area.
The thing is over time genes tend to mutate in ones body and expressors of those genes are turned on and off. NOVA had a great program on DNA and it compared the genes and the hormones that act on them.
This could explain why one identical twin is gay and one is straight. (I am just using that as an example, there is no proof so far of expressors being responsible for gay or straight people)
It showed identical twins as babies were identical and by the time you hit old age the expressors acting on those genes made the DNA much different as time goes by.
Remember not only do you have genes but there are hormones acting on those genes that turn them on and off during your life and they don’t get activated at exactly the same time. This would cause varience in twins who are identical.
And just how many of those 6 or 7 people live in the victim’s hometown and managed to draw enough suspicion that the police even bothered to test them? If you’re going to take the view that statistics can never tell you anything certain, then nothing can ever tell you anything certain, because all the knowledge we have of anything is purely statistical.
And where do you think those hormones come from, eh?
They are influenced by internal and external factors so even identical twins won’t produce the same levels of various hormones throughout their lives.
The statistical probabilities are based on analysis of a sample population. We obviously don’t have a sample of the DNA of every person in the U.S. The FBI’s CODIS database has over 4.5 million DNA profiles identifying 13 markers, so you can draw some pretty good statistical conclusions when comparing an individual’s DNA to the database.
The makeup of the database is obviously key. You probably wouldn’t be able to draw a good statistical conclusion on the DNA of a person of European ancestry if you only have 6 alleles identified and the database you’re comparing the sample to consists of 200,000 DNA samples from Hispanics. That’s why I support increased sample collection in the U.S. I think the privacy concerns are largely overblown and more samples means more accurate comparisons and a means of exonerating the falsely accused.
running calculations…
never mind