Biomedical Engineering Reference
In-Depth Information
In cataloging potential sources of errors, it's important to consider that the accuracy of protein
structure modeling is inherently limited by the accuracy and purity of the NMR and X-ray
crystallography data that form the basis of protein structure templates. Each experimental method is
associated with specific artifacts that should be considered when the results of comparative modeling
or ab initio predictions are evaluated. For example, both techniques produce erroneous data if there
is lack of homogeneity in the protein samples. In the more time-intensive X-ray crystallography,
these errors appear as a range of values instead of a single value for the distance between two
atoms. In contrast, with NMR methods, the data returned for inter-atomic distances is often a single
value that represents an average value. As such, errors in NMR data that result from populations of
proteins in the sample may be more difficult to detect. However, in many instances, choosing NMR
over X-ray crystallography or vice versa isn't possible, especially in niche areas. For example, NMR
techniques are limited to small- and medium-sized protein molecules.
