Biology Reference
In-Depth Information
detect a global sequence to structure incompatibility and errors corre-
sponding to topological differences between template and target. They
also allow the detection of more localized errors such as
-strands that
are “out of register” or buried charged residues. However, none of these
methods has allowed the detection of the more subtle structural incon-
sistencies, which are often localized in nonconserved loops, and thus can-
not provide an assessment of the correctness of their geometry.
β
5. Applications of Protein Models
The suitability of protein models for specific applications depends criti-
cally on their correctness and accuracy. There is a wide range of applica-
tions for comparative models, such as designing experiments for
site-directed mutagenesis or protein engineering, predicting ligand bind-
ing sites and docking small molecules in structure-based drug
discovery,
65,66
studying the effect of mutations and single nucleotide poly-
morphisms (SNPs),
67,68
phasing X-ray diffraction data in molecular
replacement,
69,70
as well as engineering and designing proteins.
Hereafter, we will review a number of applications of models built with
SWISS-MODEL. As will be seen, it is important to avoid any dogmatism
when considering the applicability of protein models. There are obvi-
ously some guidelines that should be considered; for instance, one
should use only high-accuracy models (
50% sequence identity between
target and template) to run
in silico
docking experiments or molecular
replacement attempts. However, these guidelines have their cohort of
exceptions, as every modeling project has its own peculiarities.
>
5.1. Functional Analysis of Proteins
The most important aim of genome sequencing projects is to assign a
biological function to the many newly discovered genes and their protein
transcripts. Insights into the 3D structure of a protein can be of great
assistance in assigning its molecular function, while its biological role and
localization are much more difficult to relate to its structure. For
instance, even if we know that a given protein is a protease and can pre-
dict its S1 pocket substrate specificity from its active site configuration,
