Protein crystallography flexibility

There are a multitude of ways in which one can abuse this approach. Of primary importance is the quality of the pharmacological data. Inclusion of a compound with a different mechanism of action should ultimately lead to a null hypothesis, i.e. no consistent pharmacophore, but may require much effort to detect. Alternatively, the procedure may be useful in identifying compounds to be further investigated pharmacologically, similar to the use of outlayers in QSAR. One objection often raised is the obvious flexibility associated with proteins. The goal is not to compete with protein crystallography, but rather to derive a functional description of the volume available for... 

This algorithm does not take into accormt the possibility of noise in the coordinates of the points. For proteins, the coordinates of atoms are approximations to a true position Proteins are flexible, with fluctuation about a mean position. Moreover, the physical experiments that provide information on the coordinates (usually X-ray crystallography and NMR spectroscopy) have a degree of experimental uncertainty. When superposing two models for the structure of one protein, the cRMS value is therefore a combination of the actual fluctuation between the two models and of the noise level contained in the two models. Noise is even more important for the superposition of two proteins of different lengths. 

Much information about flexibility of protein structure has also come from x-ray crystallography data. It is either in the form of atomic mean-square displacements... 

---