I am aware of the advances being made to crystallize membrane proteins, as that particular field was advancing when I was in graduate school in the late 1990s.
The problem with the protein I worked with was that it is incredibly unstable. I used to synthesize it in vitro with 35S labeling; even if I ran a gel immediately after synthesis, the lane was a mess of degradation products. I never quantitated how much was lost to degradation, but I can estimate that less than a quarter—probably less than 10%—of the protein was intact.
That instability makes crystallography impossible.
We were collaborating with a crystallographer to try to crystallize smaller fragments of the protein, for instance, just the basic helix-loop-helix domain—but we just couldn’t synthesize it in quantities sufficient for crystal studies. So when someone published a crystal structure of a related basic helix-loop-helix protein, we were ecstatic.
It’s a lot harder to detemine structure without being able to get crystals. you need to be more clever.
Computer modeling and force fields are very useful tool now. Back when you were doing that they quite nascent and needed expensive computers.
Did you ever figure out why your protein was so unstable?