Chemistry Reference
In-Depth Information
Using amino acid sequences (with information from other homologous
structures), homology modeling is able to generate protein structural models.
These can be built, taking into consideration similar structural stretches, and
extrapolating features from the known homologous structures. If multiple
structures are available such as template structures (basis), the procedure can be
better performed. The process involves identifying structural templates, sequence
alignment, modeling of structurally conserved regions, loop modeling, side chain
modeling, model optimization and model validation.
The first step is to search for proteins with known 3d structures. These should be
related to the target-protein (target sequence). Function similarity, structural
knowledge, evolutionary correlation and expression by the same group of genes
should be taken into consideration.
Since sequences of amino acids of proteins of unknown structures can be
obtained, sequence similarity is often used to search for homologous proteins. In
this approach, a number of requirements are necessary to determine the degree of
sequence identity (construction of target structure model). It is necessary to have
alignment with respect to one or more sequences of known 3D structures as well
as prediction of the secondary structures that the amino acids of the target
sequence will assume.
The first task is to search for proteins with known 3d structures that could be used
as templates for modeling the unknown structure. They should correlate with the
target protein. The Basic Local Alignment Search Tool is a useful homology
software. It searches the sequence database (aligns the query sequence with each
sequence in the database). It uses an heuristic algorithm to search for high scoring
of sequence alignments. Specific tools are used to assess the degree of similarity
between fragments of the sequences (local alignment). This is to identify, from a
biological/structural point of view, important similarities. PDB and other
databases can be used. Algorithms for sequence similarity searches can be applied
Once the template structures have been selected, one proceeds to align the target
with the template sequences. If the percentage identity between compared
sequences is high, the correct alignment should proceed. However, in the cases of
