Secondary and Tertiary Structure Prediction of Proteins: A Bioinformatic Approach - Complex System Modelling and Control Through Intelligent Soft Computations - page 551

7 Applications of Protein Structure Prediction

Homology/Comparative modeling plays an essential role in structure based drug
design. For example representative structures produced by in silico screening forms
the basis of generation of three-dimensional structures of the remaining proteins
encoded in the various genomes that can be predicted by homology modeling
(Takeda-Shitaka et al. 2004 ). Comparative modeled proteins may be used for
predicting the binding modes and affinities of different drug compounds as they
interact with protein binding sites in structure-based drug design. Computational
approach to this problem is usually termed as molecular docking. The goal of
ligand-protein docking is to predict the predominant binding mode(s) of a ligand
with a protein of known three-dimensional structure. Docking can be used to
perform virtual screening on large libraries of compounds, rank the results, and
propose structural hypotheses of how the ligands inhibit the target (Morris and Lim-
Wilby 2008 ). However, it is widely accepted that docking with comparative models
is more challenging and less successful than docking with crystallographic struc-
tures. Comparative models are not only useful in protein-ligand, but also useful in
protein-protein docking (Vakser 1997 ).
Comparative models can also be used for testing and improving sequence
structure alignment (Wolf et al. 1998 ). Based on the alignment of known structures,
alignments can be well de ned even for a new target sequence. Apart from the

presence of functional motifs or the signature sequences, calculated electrostatic
potential around the protein structure may help in predicting the protein function
(Drew et al. 2011 ).
Protein models by comparative method can be also used to decipher important
residues for biological activity as well as function of the protein. These models can
be helpful in designing mutants to test hypotheses about protein functions (Boissel
et al. 1993 ). On the basis of its primary sequence and the location of its disul

de
bonds, erythropoietic hormone erythropoietin was modeled by homology modeling
which predicts a four alpha-helical bundle motif, in common with other cytokines.
Deletions of 5 8 residues from erythropoietin hormone erythropoietin protein

within predicted alpha-helices resulted in the failure of export of the mutant protein
from the cell (Boissel et al. 1993 ).
Comparative models can also be used to explore the substrate speci

-

city in
several enzymes. After the crystallization of the bacterial leucine transporter protein
LeuT, development of 3-D computational models were used for structure-function
studies on the plasmalemmal monoamine transporters (MATs). LeuT-based MAT
models were used to guide elucidation of substrate and inhibitor binding pockets.
Moreover, molecular dynamics simulations using these models provided insight
into the conformations involved in the substrate translocation cycle (Manepalli et al.
2012 ).
Comparative models have been used in conjunction with virtual screening to
successfully identify novel inhibitors over the past few years. Novel inhibitors of
dihydrofolate reductase in Typnosoma. cruzi (the parasite that causes Chagas