Secondary and Tertiary Structure Prediction of Proteins: A Bioinformatic Approach - Complex System Modelling and Control Through Intelligent Soft Computations - page 542

decreases with the number and length of gaps in the alignment. Multiple sequence
alignment by various freely available programmes e.g. ClustalW (Larkin et al.
2007 ) Mufft (Katoh et al. 2002 ), Kalign (Lassmann and Sonnhammer 2005 ),
Probcons (Do et al. 2005 ) etc. and a development of phylogenetic tree by freely
available programmes e.g. MEGA (Tamura et al. 2013 ) and PHYLIP etc. can help
in selecting the template from the subfamily that is closest to the target sequence.
HHpred ( http://toolkit.tuebingen.mpg.de/hhpred ) is one of the best servers which
the can even detect very distant relationships between the target sequence and the
solved PDB structures signi

cantly. This is the first server that is based on the

pairwise comparison of pro le Hidden Markov Models (HMMs) (S ö ding et al.

2005 ).
The similarity between the of the template and the environment in

which the target needs to be modeled should also be considered. The quality of the
experimentally determined structure is another important factor in template selec-
tion whereby high resolution X-ray crystal structure is more preferred for template
selection than that of low resolution crystal structure. Multiple templates rather than
selecting a single template, generally increases the model accuracy. A good protein
structure model depends on alignment between the target and template.

' environment '

3.3.2 Alignment Correction in Core Regions

An accurate alignment can be calculated automatically using standard sequence-
sequence alignment methods, for example, Blast2seq ( http://blast.ncbi.nlm.nih.gov/
Blast.cgi ) and dynamic programming based Needle global sequence alignment
( https://www.ebi.ac.uk/Tools/psa/emboss_needle/ ). In low sequence identity cases,
the alignment accuracy is the most important factor which affects the quality of the
predicted model. Alignments can be improved by including structural information
from the template protein structure. Gaps should be avoided in core regions mainly
in secondary structure elements (which are found to be conserved in most cases),
buried regions and between two residues that are far in space. It is important to
inspect and edit the alignment manually by many tools e.g. Bioedit ( www.mbio.
ncsu.edu/bioedit/bioedit.html ) etc., especially if the target-template sequence iden-
tity is low.

3.3.3 Backbone, Loop and Side-Chain Modeling

Creating the backbone is essential for modeled protein structure. For backbone, we
simply copy the coordinates of those template residues that show up in the align-
ment with the model sequence. If two aligned residues differ, only the backbone
coordinates (N, C

, C and O) can be copied. If they are the same, we can also
include the coordinates of side chain amino acid residues.

α