I'd add that you need to do the homework first: If you can adequately formulate and articulate the problem that the lab is solving, and that one cool trick you have that can solve it, that's already a great signal that you would work well (Chances are your trick won't work, but that's only because biology is merciless).