Admittedly the Autism example is not the best, however I believe that it does a decent job of getting the point across. The studies linking autism to Atheism are not of the highest quality, however a clear correlation has been demonstrated. Additionally, there is theory as to why high functioning autistic people would be more likely to be atheists than neurotypical people. Having a theory and having demonstrated a correlation is at least supporting evidence that Autism is causative of Atheism, but I agree that currently the research is not capable of proving causation. I do believe that a more sophisticated study will demonstrate that Autism has a causative relationship with Atheism. As far as citations go, here are some links and relevant snippets from them: http://blogs.discovermagazine.com/gnxp/2011/09/atheism-as-mental-deviance/ http://www.scienceonreligion.org/index.php/news-research/research-updates/490-autism-and-ir-religiosity I mean, it just seems pretty obvious to me that if points A and B are true, that point C must follow. The probability of the guess being correct is dependent on the degree of correlation between the presence of the gene and the manifestation of autism, as well as the degree of correlation between the presence of autism and atheism. We are not yet capable of diagnosing Autism via genetic analysis alone, however certain genes increase the risk of autism to several times greater than that of the general population. Additionally, further research needs to be done to establish a causative link between Autism and Atheism, however, current research is supportive of such a link existing. So I agree that today we might not be quite to the point that we can make highly accurate guesses as to a persons belief in God based upon the methodology I suggested. However, I do believe that we can currently guess, with higher than random accuracy, if a person is an Atheist or not, by by using the methodology I suggested. Even an increase in accuracy of half a percent would indicate that analysis of DNA can be used to predict a persons belief (or lack thereof) in God, with better accuracy than can be obtained without the DNA analysis (in which case the best guess would be that the persons probability of Atheism is in line with that of the general population). As further research is carried out on the link between Autism and Atheism as well as the link between the presence of certain genetic mutations and the presence of Autism, I strongly believe that our predictions will only increase in accuracy. How is this not what I am talking about? See the links I provided. : I believe that I have made a pretty reasonable case, but I can see that as far as Autism and Atheism goes, I could have selected a better example with more research available. How about this study: To me this seems to pretty clearly state that gene variation can be used to accurately predict the behaviors of a person playing an economic game.