Protein crystals have often been likened, by some well respected experts within the field, to a block of "swiss cheese". Not only does a protein crystal have an enormously high solvent content, the atoms which comprise the unit cell diffract X-rays rather weakly . If this were not bad enough, the unit cells are sometimes not arranged periodically. Now add to this a few holes,impurities and crystal imperfections and you suddenly get something that resembles a nightmare to study - or like I mentioned before 'swiss cheese'. That's if you even lucky enough to form a crystal in the first place.
You may possibly be wondering why all this is important. Protein crystals are important because the main method used to determine large protein structures is via X-ray crystallography (see below) .In order to perform X-ray crystallography you need a good quality protein crystal that diffracts X-rays well. The production of good quality protein crystals IS the major bottleneck in protein structure determination. Large amounts of effort are dedicated to this task and the process of protein crystallization is still poorly understood. As mentioned previously ,before you find the structure of a protein ,which will largely dictate protein function , we must first create crystals .This process is currently trial and error.