I don't think it is really different at all. For example: http://www.telegraph.co.uk/science/science-news/5270316/Anger-is-in-the-genes.html This study shows that Anger and violence definitely correlate. So do you really think that somebody who doesn't have the "TT" or "TC" version of the studied gene, is as likely to commit a violent crime as somebody who does? It was never claimed that DNA can tell whether someone is about to commit a crime or not. The only claim I made was that DNA can be used to determine the probability that someone will EVER commit a certain type of crime, and that fingerprints cannot be used for this. DNA can also be used to predict with better than random accuracy if a person believes in God or not, if a person is a Democrat or a Republican, and all kinds of other things. So DNA contains a lot more information of interest than a fingerprint does. Even if genes are only markers of possibility, they can still be used to increase the accuracy of a guess to greater than that of the general population. It is really simple statistics. If someone in the general population has a 1% chance of being atheist that means 1 out of 100 people are atheists. If people with a certain gene have a 90% chance of being autistic, that means 90 out of 100 people with that gene are autistic. If autistic people have a 90% chance of being atheists, then 90 out of 100 people with that gene have a 90% chance of being atheist.