Page 263 - Big Data Analytics for Intelligent Healthcare Management

256 CHAPTER 10 COMPUTATIONAL BIOLOGY APPROACH ON GENETIC
DISORDER

After assimilating the structures of the ten selected proteins from the PDB database and the list of all
the corresponding drugs from the drug bank, each protein was thereafter interacted (docked) with all its
drugs through virtual screening of AutoDock software. By observing the binding energies of all the
interactions obtained through virtual screening, the best interactions can be selected and the drugs inter-
acting with the highest or lowest binding energies are identified as the best drugs for AD. Auto dock is
an important tool capable of accurately predicting the different conformation and binding energies of
ligands with macromolecular targets.

10.6 DIFFERENT ALGORITHMS RELATED TO DOCKING
Docking methods are classified into different categories depending on their search algorithms, which
are explained by a set of parameters and rules to analyze the conformations. The docking algorithms are
classified into two major groups: one is flexible docking and another is rigid docking. In the flexible
docking algorithm, the receptor and ligand are flexible or freely identify the conformation binding side
whereas the rigid docking algorithm does not consider the flexibility of the receptor or ligand. In the
rigid docking algorithm, the ligand flexibilities are not capable of identifying binding sites for several
proteins. Search algorithms are required to predict the conformations in the docking application that are
defined by set of parameters and rules. In rigid docking, geometrical matching between two molecules
is needed [49]. For identification of the ligand binding site of different proteins, the flexibility is not
taken into account in rigid docking. Generally, the root mean square deviation (RMSD) is calculated
between two sets of atomic coordinates. The RMSD is defined between two sets of atomic coordinates,
calculated from the following equation:

v ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u N
1
u X 2 2 2
ð
RMSD ¼ t ð x ci x di Þ + y ci y di Þ + z ci z di Þ
ð
N
i¼1
The x ci , y ci ,and z ci were indicated as center and x di , y di ,and z di indicated dimension of the three dimen-
sional structure. The crystallographic and other one for the coordinates obtained from docking simulation
which is taken over all N atoms. Normally in docking simulations, the RMSD values should be less than
˚
1.5A [50]. Both intra and inter molecular distances are expressed using distance geometry algorithm
methods. These molecular distances are used for prediction of structures or conformations. Different
algorithms are used in order to make several binding conformations between the ligand and receptor.
(a) Shape matching algorithm (SMA)
The shape matching algorithms are methods that consider the structural overlap between ligand and
receptor. This approach may discover conformational binding sites of protein using a macromolecular
surface search.

(b) Simulated annealing algorithm (SAA)
The simulated annealing algorithm is used for prediction of protein structure studies as well as con-
formational analysis. In this method, the entire bimolecular system undergoes a specific kind of dy-
namic simulation. The temperature is decreased gradually at regular intervals of time.

(c) Genetic algorithm (GA) and distance geometry (DG) algorithms