Secondary and Tertiary Structure Prediction of Proteins: A Bioinformatic Approach - Complex System Modelling and Control Through Intelligent Soft Computations - page 544

www.gromacs.org/ ) , AMBER ( www.amber.scripps.edu ), and CHARMM ( http://
www.charmm.org/ ) which are freely as well commercially available for protein
model re

nement by correcting the overall protein structural geometry. One of the
recently developed re ne is computationally inex-

pensive and consumes only few minutes of CPU time to re

nement methods called 3Dre

ne a protein of typical
length of 300 amino acid residues (Bhattacharya and Cheng 2013 ).

3.3.5 Model Evaluation or Validation

The predicted model must be checked for

(a) Errors or distortion in side chain packing of the modeled structure.
(b) Distortions or shifts in correctly aligned region of target with the template
structures.
(c) Distortions or shifts of a region that does not align with any of the template
structures.
(d) Distortions or shifts of a region that is aligned incorrectly with the template
structures.

Structural model accuracy is mainly based on global distance test (GDT), which
is an average percentage of model C α atoms within a specified distance threshold to
actual native conformation (Jauch et al. 2007 ).

1
4 ð

max 1A 0 max 2A 0 max 4A 0 max 8A 0 GDT ¼ þ þ þ Þ ð 1 Þ

Equation 1 GDT score where C n 0 is the number of atom pairs closer than distance

of n = 1, 2, 4 and 8A 0 .
TM score is another method for validating the model accuracy to score the
topological similarity of target and template structures, where the score near to 1.00
is the best predicted near-native model against the actual experimental structure for
a target (Xu and Zhang 2010 ). MaxSub is another new and independently devel-
oped method which aims at identifying the largest subset of C(alpha) atoms of a
model that superimpose

' '

over the experimental structure, and produces a
single normalized score that represents the quality of the model (Siew et al. 2000 ).
Various programmes and web-servers are available for checking the quality of
the model. One of these is Procheck (Laskowski et al. 1993 ) that generates the
Ramachandran Plot, which illustrates the stereo chemical quality of the protein
model. Few popular web servers for protein structure quality validation and their
description are listed in Table 5 .
Few popular web servers for modeling the protein structure by homology or
comparative modeling method along with the webpage URL and description are
listed in Table 6 .

well