For example, determination of disorder from crystal structures can be particularly problematic for two reasons:
inconsistent methods of documentation in PDB archive entries, and
the difficulty in distinguishing wobbly domains from true disorder.
Disorder assignment by CD can be particularly problematic, due to difficulties in resolving contributions of different secondary structure types combined with the lack of position specific information. The existence of structured proteins that contain no traditional secondary structure and therefore give a far UV spectrum similar to a completely disordered protein requires that CD spectral evidence be corroborated by other studies (such as proteolytic degradation rate measurements) before protein can be confidently classified as disordered. Even disorder determined by NMR can have shot comings as discussed previously (Garner et al. 1998).