a structural biology technique used to figure out the atomic or near-atomic structures of molecules and/or molecular complexes (groups of molecules bound to each other). Basically, the goal is to determine where molecules are located relative to one another in 3-D. It’s often used to “solve” the structures of small molecules like drugs as well as bigger “macromolecules” including proteins and nucleic acids (DNA or RNA). It works (or at least you hope it’ll work – it can take a lot of screening…) by convincing a ton of copies of the molecule to crystallize (freeze in an orderly arrangement). Then you beam x-rays at that crystal. When those x-rays interact with the atoms in the crystallized molecules they get kinda “bounced-off” (scattered). And then those scattered x-ray beams interact with one another – some cancel each other out, others add together, and, thanks to the orderly crystalline arrangement, if you measure when the rays hit a detector, you end up with a series of dots with different intensities. This is the “diffraction pattern.” Then you take this diffraction pattern and “work backwards” to figure out where the atoms were that did the scattering.
Crystallography can give you really high-resolution (minimally-fuzzy) information about molecular structures, including where drugs bind. But it requires the molecules to crystallize which a lot of proteins aren’t very happy with, so you often have to make changes to the protein such as removing loopy bits in order to get it to stop flopping around! And then you have to screen a bunch of crystallization conditions… sometimes a bunch a bunch. Lot’s more in the linked post. When it comes to proteins, chances of success are better with small-medium sized proteins.
We have Dorothy Crowfoot Hodgkin to thank for much of macromolecular crystallography, so check out that post too.