> experimentally solve multiple high-resolution 3D models per week I thought this was only true if you already have a structure; otherwise you typically run in

> experimentally solve multiple high-resolution 3D models per week I thought this was only true if you already have a structure; otherwise you typically run into the [phase problem](https://en.wikipedia.org/wiki/Phase_problem https://en.wikipedia.org/wiki/Phase_problem), which is often a significant hurdle. But I haven't done much biochemistry in years, so there might be better approaches than MAD/MIR/etc. that make the phase problem a non-issue.