One of the things worth pointing out is that biological effects are hilariously noisy. Even many of our best genetic correlates aren't that clear cut. A lot of

One of the things worth pointing out is that biological effects are hilariously noisy. Even many of our best genetic correlates aren't that clear cut. A lot of what we know still relies on large sample sizes to help cut through real confounding factors. And if that's what it takes to detect causal genetic effects, then we'll have to understand whatever we do to 'fix' these (whatever approach we take) will have to deal with the same type of issues. In other words, personalized medicine isn't going to be this amazing thing that will let us conquer all our illnesses. It will be a tool that will bring significant improvements, possibly even order of magnitude improvements to those with the right genes or illnesses, but it's not magic.