Simulation and 3 d visualization of complex molecular structure for study of protein and nano materials

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 04 Issue: 02 | Feb-2015, Available @ http://www.ijret.org 660
SIMULATION AND 3D VISUALIZATION OF COMPLEX MOLECULAR
STRUCTURE FOR STUDY OF PROTEIN AND NANO MATERIALS
Himanshu S Mazumdar1
, Shivangi R Shah2
1
Professor and Head, Research and development Center, Dharmsinh Desai University, Gujarat, India
2
Student, Information and Technology, Dharmsinh Desai University, Gujarat, India
Abstract
Simulation and visualization of complex bio-molecules are gaining importance; this is because of the fact that functional
properties of such molecules are more dependent on their 3D structures. One of the challenges of computational biology is
prediction of structures of protein from amino sequence. Three dimensional visualization of molecular structure is also an
important research goal in nano engineering. Such visualization, simulation and animation of reaction dynamics are essential for
modern chemistry. Force field simulation of large number of atoms in complex molecular structure is computationally
challenging. To find equilibrium of force fields and resultant dynamic structure of large atomic clusters which assembles to
predictable molecules is one of the major goals of computational biology. In this paper we describe a simple but efficient
algorithm (MoliSim3D) to simulate and dynamically visualize in 3D the assembly of atoms in presence of internal and external
forces. It helps monitoring and measuring different bond angles, dimensions and displacements of selected portions from
thousands of atoms forming target molecule. Pattern recognition, pattern error detection and reaction control are advance tools
of virtual molecular assembly. The proposed system is basically intended as educational tool and to help researchers in
understanding complex molecular dynamics and functionality with high degree of confidence in simulated environment.
Keywords: Simulation, 3D Visualization, Structure, Dynamic
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
In the field of molecular biology, protein is the major
structural constituent. It is the protein that represents most of
the complex life functions. Major constituent of protein are
amino acids which are most fundamental molecule in life
present on the earth. Research has been carried out using
complex searching algorithms for efficient usage of large
Proteins Data-Bases (PDB) [5], similar protein search in
large PDB like UniRef100 [6], protein structure prediction
[9] and multiple protein sequence alignment [10]. Two
major facts for protein are: (a) Different Protein structure is
the cause for different protein function [3], (b) Different
Protein function depends on different physical and chemical
parameter, so it can be said that a given type of protein
sequence can form multiple structures and those structures
further can provide different functions. Enzymes are protein
catalysts which can perform most complex chemical
function by making or breaking the chemical bonds [1]. A
protein needs to be positioned accurately in 3D space for
efficient reaction in a given substrate. It is scaffolding that
tethers other proteins to form stable complexes structure by
implementing accurate docking [7]. For a large organic
molecule the atom level representation using computer
graphics is a complex task. From atom level representation
of complex molecule, it is difficult to conceive the
underlying pattern which needs to be identified for probable
efficient tethering. Shape and structure of a protein causes
different function. A function further uses 3D structure more
than the sequence. Even though DNA molecule contains
necessary information for life, the important process like
maintenance, replication, defense and reproduction are
carried out by proteins. It is the chemical and physical
structures of a protein that determines its activity [4].
Each structure basically contains thousands of atoms per
molecule. For this reason, it is often consider convenient to
simplify the problem by using a hierarchical description of
protein structure in which successive layer of the hierarchy
describe increasingly more complex levels of
organization[8]. Precision of structure is important along
with the atomic detail as a small alteration in the atomic
composition may dramatically change the functionality like
hemoglobin to sequel shell structure which defers by Amino
acid sequence change from glutamic acid to valine in 2 beta
sheet out of 4 chains of hemoglobin [2]. Characterization of
natural proteins and altering its function require not only the
knowledge of its chemical composition but its physical
structure and orientation in different dynamic conditions.
Dynamic force field simulation of such molecules allows a
researcher to visualize and plan desired modification
strategies of controlled reaction. Such simulations will
reduce the trial and error experimentations in determining
structure of protein and design of targeted drugs. This
process is further accelerated with availability of database of
structure of known targeted drugs. Three dimensional
patterns search for complementary structures in database
automates the drug design process. This paper describes an
educational or research tool that can perform the task of
simple chemical lab experiments. One of the novel features
of the proposed system is motion dynamics. This is
implemented using an iterative algorithm of inter-atomic
force equilibrium difference-equation. The algorithm is
validated by measuring the resultant bond angles of
synthesized well known molecules and standard amino acids
as shown in figure-1.

_______________________________________________________________________________________
Fig-1: Results of dynamic force field equilibrium condition
of trigonal amine NH3 bonds in (a) and tetrahedron bonds of
methane CH4 in (b).
2. SYSTEM ARCHITECTURE
The architecture of MoliSim3D consists of a 3D graphics
editor, inter-atomic force simulator, amino acid database and
powerful utilities for 3D viewing of atoms, bonds and
molecules. Most of the activities of the simulator are around
the atom-bond 3D position database as shown in figure-2
and described in section-3. System is organized based on
different menu driven utilities for composing, viewing,
simulating reaction, database management and
documentation. Important features of MoliSim3D are
explained in section-4 and Section-5. The simulator
generates a script file consisting of list of atoms with its
attributes, list of bonds between atoms and current 3D
position of all the atoms as shown in figure. During
simulation atom-bonds and atom positions are continuously
updated. This script file is saved as a library component of
molecules. Number of these components can be loaded
together and allowed to react to create bigger molecules like
proteins.
Fig-2: System architecture of MoliSim3D. All the activities
of IO, GUI and force simulations are around the atoms,
bonds and 3D position database of atom descriptor array.
3. DATA STRUCTURE
The heart of the MoliSim3D is atom-bond 3D position
database as shown in figure-2. This database consists of
array of atoms with properties and array of connections.
Attributes associated with each atom are explained below.
3.1 Atom Structure
Most activities of this simulator architecture are around the
array of atom structure which confirms acceptable
nomenclatures of the molecular chemistry. Important
elements of the atom structure are described below.

_______________________________________________________________________________________
Fig-3: Dry Chemistry Lab application screenshot. On the left controls are provided for molecule composition and animation of
force equilibrium. On the right script is generated describing the assembled molecules. The 3D display in middle shows animation
of molecular dynamics and structure formation.
3.2 Atom Number
This field is used as unique atom ID. This ID is displayed on
atom body and used in molecules descriptor list to define
bonds and atom's 3D locations. Atom number is reorganized
in run time after any new inter molecular bonding.
3.3 Atom Symbol
Atoms are labeled as 'H', 'O', 'C', 'N', 'S', and 'L'. These
labels are associated with atom's color, diameter and valance
used in molecules descriptor list.
3.4 Atom Color
Atom color is associated with the atom symbol and is
automatically updated. This parameter is used for 3D atom
and bond color display.
3.5 Atom Location
This field is used to hold the current position of the
associated atom using normalized 3D coordinates. This is
the key parameter for atomic and molecular reaction.
Continuous updating of this parameter results in 3D
animation of molecular reaction and structure visualization.
3.6 Atom Select
This is a Boolean variable used mostly for interactive
editing and visualization of molecule. Multiple auto-
selections such as group select help identifying and
relocating connected atoms. Selected atoms are visually
highlighted and used for adding, removing, changing or
moving atoms.
3.7 Atom Source Link
This is an integer array field that contains list of atom IDs
that are greater than this atom's ID and connected to this
atom. Collection of atoms bonding list defines the bonding
details of the molecule or connected atoms group. No bond
is duplicated in these lists for a molecule.
3.8 Atom All Link
This is an integer array field that contains list of all atom
IDs that are connected to this atom. All Link array of an
atom contains all bonding details of source and destination
atom IDs in a given molecule or connected atoms group.
All bonds are duplicated in these lists. These arrays are used
to calculate the dynamic force fields and continuously
updated during reaction simulation.
3.9 Atom Diameter
This field provides the relative diameter value of an atom
normalized with respect to diameter of Hydrogen atom. This
parameter is used for realistic 3D projection of molecules.

_______________________________________________________________________________________
4. COMPOSING AND EDITING FEATURES
As shown in figure-4 creation of a new 3D molecular
structure or modifying an existing molecule is performed
using a set of powerful but simple editing commands. The
editing operation is performed in X-Y, X-Z and Y-Z
projected plane of 3D space. Following are brief
descriptions of important commands.
4.1 Add Atom
A new selected type of atom can be placed on any of the 3
plains using mouse double click. After placing it in a given
plain it can be seen in other plane and dragged to desired
position in 3D space. Any new placement updates the atom
and location entries of source list of text window at the
right.
4.2 Add Bond
Any two atoms can be bonded together using mouse click.
Any editing actions are immediately reflected in the source
list of the target molecule.
Fig-4: Screenshots of 3D molecule composition. (a) image
shows ‘Add Atom’ and ‘Add Bond’ operations. On (b),
editing utilities like ‘Remove atom’, ‘Remove bond’ and
‘Change atom’ are shown.
4.3 Remove Atom
Any atom can be removed from screen and list by double
clicking on it if Remove Atom radio button is selected.
4.4 Remove Bond
Any bond can be removed from screen and list by double
clicking on center of the desired bond if Remove Bond radio
button is selected.
4.5 Change Atom
Type of any atom can be changed in screen and list by
double clicking on it if Change Atom radio button is
selected along with selected appropriate destination atom
type.
4.6 Lone Pair
Insertion of Lone Pair is performed by adding 2 Lone pair
elements and three bonds connecting them with associated
atom. There is an external control to trim to calibrate the
bond angles.
5. VISUALIZATION OPTIONS
Composed molecules are projected in 2D plane. A set of
utilities are provided to enhance the viewing precipitation.
Interactive measurement of 3D bond angles is an important
feature of visualization. Descriptions of important viewing
options are given below with relevant screenshots in figure-
5.
5.1 Offset
Displacement of the 3D position of a selected atom, group
of connected atoms or all atoms in reference to an origin or
to a previous position is performed interactively using
mouse drag. A complete 3D offset is a two step operation
performed in two different planes.
5.2 Rotate
Rotation in 3D space of selected group in reference to group
center or all atoms in reference to center of image are
performed in X or Y axis using mouse drag.
5.3 Spin
This operation invokes continuous rotation of selected group
in reference to group center or all atoms in reference to
center of image in Z axis. Visualization of reaction or
angular stabilization is enhanced in this mode.
5.4 Place at Center
This operation repositions average center of all atoms or
selected atoms to the center of the screen. This operation is
useful when atoms are drifted away from center under the
influence of force fields.

_______________________________________________________________________________________
5.5 Measure Bond Angles
Interactive measurement of 3D bond angles between any
two bonds having a common node atom is performed by
pointing connected atoms in sequence. Result in degrees is
displayed in a text box as shown in figure-1.
Fig-5: Screenshot of multiple molecules being composed. In
(a), 3 molecules are assembled using manual editing and in
(b), force field equilibrium is applied to get stable structures.
6. MOTION DYNAMICS
Simulation of motion dynamics is one of the important
features of this simulator. Motion dynamics in MoliSim3D
consist of computation and visualization of the temporal
structure of atomic bonds forming molecule. Initially the
atoms are placed using 3D graphics editor and are connected
according to the chemical formula using bonds. Force
equilibrium algorithm is applied to the assembled structure
to compute the 3D co-ordinates of each atom in a given
molecule under inter-atomic forces. The algorithm is
iterative and uses difference equation to compute the atom
co-ordinates. The intermediate co-ordinates of the atoms,
when visualized produce effect of visual simulation of
structure stabilization. There are 4 utilities associated with
motion dynamics as described below.
6.1 Randomize Atom Position
The initial condition of atoms could be randomized using
randomize utility. This utility helps to examine different
possible stable states of any assembled molecule.
6.2 Stretch Amino and Carboxyl Ends
A special utility for amino acid or protein is included to
generate an artificial repulsive force field between free ends
of amino and carboxyl group atoms. This allows viewing of
the entire amino acid chain of protein as a linear primary
structure.
6.3 Force Equilibrium Algorithm
A computationally simple but effective algorithm is
developed to organize large number of bonded atoms in 3D
space very similar to their natural 3D structure in dynamic
equilibrium for animation and final static position. This is
further enhanced using chemical bonding rules to simulate
molecular reaction like peptide bond formation from random
initial states. A simplified flowchart of force equilibrium
algorithm is shown in figure-6. It works on condition for
attraction and repulsion forces where only end atoms are
repulsive to each other. Bond angles in the stable or quasi-
stable state are measured and are in agreement with
experimental results as shown in table 1.
Fig-6: Flowchart of Force Equilibrium algorithm. The push
and pull actions are implemented using difference equation
of molecular dynamics.
6.4 Chemical Reaction Algorithm
Animation of amino acid reaction utility is implemented
under force field equilibrium. This algorithm uses rule based
interaction between amino and carboxyl group of nearest

_______________________________________________________________________________________
molecules. This animated action produces peptide bond and
releases H2O molecule as illustrated in screenshots of figure
11 and 12. A simplified representation of the reaction
algorithm is shown in figure-7.
Fig-7: Flowchart of chemical reaction between standard
amino acids.
7. PERFORMANCE ANALYSIS AND RESULTS
Different test are performed to validate the accuracy and
performance of the proposed system. Actual angle and
simulated angle between atomic bonds are compared after
force field equilibrium. Average test results of 10
experiments are shown in table-1.
System is also tested with 100 randomly selected amino
acids for dynamic equilibrium and reaction to make a single
chain of protein. The operation consists of creation of 99
peptide bonds and separation of 99 H2O molecules in pc
based desktop environment. Result indicating total 10
numbers of experiments each for randomly selecting amino
acids at constant force equilibrium of 100 is shown in table-
2. These experiments are carried out using:
 Processor: I5, operating system: windows 8.1(64
bit), Ram: 8 GB and graphics card: 1GB.
Table 1: shows the comparison of actual angle and
simulated angle between atomic bonds selected from a
molecule. Result is calculated based on average of randomly
selected bonds of 10 randomly selected molecules.
Total
no of
bonds
Molecular
geometry
Actual
bond
angle
Measured
average bond
angles of 10
samples
2 Linear 180 180.0
3 Triagonal planer 120 120.0
4 Tetrahedral 109.5 109.4
5 Trigonal
bipyramidal
90,120,
180
90.2,119.6,
179.5
6 Octahedral 90, 180 90.5, 180.0
Table 2: shows time taken and memory usage to create peptide bonds between randomly selected amino acids using force field
equilibrium, stretch and reaction utilities.
Total
no of
amino
acid
Total
no of
atoms
present
Time taken
to form
peptide(s)
Memory
taken to
form PB
(GB)
Maximum
cpu
usage(%)
2 46 3 1.1 25
5 210 5 1.1 25
10 563 21 1.2 25
20 1030 49 1.3 35
40 2060 200 1.4 35
50 1276 265 1.5 39
60 3090 595 1.5 35
100 2575 719 1.9 40
200 5150 3700 2.2 50
7.1 Display Type
Large number of different types of molecular assemblies are composed and structured under force field to test performance of the
system. All the possible outputs of different compositions are found very close to the desired structure. Some of the samples are
shown in figure- 8, 9 and 10

_______________________________________________________________________________________
Fig-8: Screenshot of standard 20 Amino Acids
Fig-9: Assembly of three nano structures: Nano sheet, Bucky ball and Nano tube

_______________________________________________________________________________________
Fig-10: CO2, NH3, CH4, NH5, SH6 and SH8 molecules being composed at (a) and 3D view of each after equilibrium at (b) with
appropriate bond angles.
7.2 Results of Molecular Reaction Simulation
Large number of amino acids are loaded and allowed to stabilize and react under force field and reaction utility. The system
completes the reaction in few minutes using the hardware configuration as mentioned earlier. A screenshot of 100 amino acids
before and after completion of reaction is shown in figure- 11 and 12. The operation took 8.5 minutes to compete.

_______________________________________________________________________________________
Fig-11: 100 amino acids consisting of randomly chosen from 20 standard amino acids are loaded from amino acid library.
Fig-12: The view while reaction of 100 amino acids as shown in figure-11. New 100 molecules include 99 water molecules along
with the linear protein is seen under stretched condition.

_______________________________________________________________________________________
8. CONCLUSION
A desktop application for simulation and 3D visualization of
molecular reaction is developed using C# Visual studio. The
application is validated using reaction of 20 standard amino
acids that are randomly placed and are allowed to interact
chemically to produce a protein chain as shown in figure 11
and 12. The system is capable of interactively composing
simple molecules like H2O with electron lone pair in CAD
like environment and supports measurement of the bond
angles in equilibrium condition. It supports large external
and internal database as input. It can load amino sequence
from entire protein database of uniref100 or from native
database of 20 amino acids. It can assemble both proteins
and other inorganic molecules. Multiple molecules could be
loaded at the same time to simulate and visualize
intermolecular reaction dynamics. One of the useful features
for portability is use of text representation of the current
state of atoms, bonds and positions in 3D with details of
bond strength and lone pairs. Its force field simulator with
randomize initialization helps exploring number of possible
stable states in equilibrium.
ACKNOWLEDGEMENTS
This work is carried out in R&D Center, Dharmsinh Desai
University Nadiad and with resources of PLANEX, PRL
project as M.Tech Dissertation work. We sincerely thank the
staff of R&D Center for providing technical support.
REFERENCES
[1] Alberts B, Johnson A, Lewis J, et al. New York
“Molecular Biology of the Cell. 4th edition”
Garland Science; 2002.
[2] Julie Ann Buford. "Sickle Cell Hemoglobin and
Malaria: An Adaptive Study of Natural Selection on
an Infectious Disease." (2004).
[3] Charles W. O'Donnell, Mieszko Lis, Jérôme
Waldispühl & Srinivas Devadas “Predicting Protein
and RNA Structures and Interactions using
Biophysically-Motivated Energy Functions and
Machine Learning” Computer Science and Artificial
Intelligence Laboratory (CSAIL)-2007
[4] Gruber, Arthur, Alan M. Durham, Chuong Huynh,
and Hernando A. del Portillo. "Bioinformatics in
tropical disease research." (2008).
[5] Maulika S Patel and Himanshu S Mazumdar,
“Utilities for Efficient Usage of Large Biological
Database,” Phil. WCECS 2010 Vol II, October 20-
22, 2010, San Francisco, USA.
“Similarity Search Using Pre-Search in UniRef100
Database,” International Journal of Hybrid
Information Technology Vol. 4 No. 3, July, 2011
[7] Catherine Qiurong Pan, Marius Sudol, Michael
Sheetz and Boon Chuan Low “Modularity and
functional plasticity of scaffold proteins as p(l)
acemakers in cell signaling” Cellular Signalling-
2012
[8] Tverdislov, Vsevolod A. "Chirality as an Instrument
of Stratification of Hierarchical Systems in Animate
and Inanimate Nature." arXiv preprint
arXiv:1212.1677 (2012).
“Knowledge base and neural network approach for
protein secondary structure prdiction,” Journal of
Theoritical Biology 2014
[10] Himanshu S Mazumdar Ankita C Baravaliya and
Maulika S Patel, “Keyword based Iterative Approch
to multiple Sequence Alignment,” Int. J. Pure App.
Biosci. 2 (3): 139-144 (2014)
BIOGRAPHIES
Himanshu S Mazumdar, Ph.D., Senior
Member, IEEE is currently working as
Professor, EC and Head of Research and
Development Center at Dharmsinh
Desai University. He has worked in
important space missions in University
College London, NASA’s Space Shuttle and Indian Space
Research Organization.
Shivangi Shah received the Bachelor
degree in Computer engineering in 2013
and currently pursuing Masters in
Information Technology and has carried
out dissertation work in Research and
Development Center during 2014-2015
from Dharmsinh Desai University

Simulation and 3 d visualization of complex molecular structure for study of protein and nano materials

More Related Content

Similar to Simulation and 3 d visualization of complex molecular structure for study of protein and nano materials (20)

More from eSAT Journals (20)

Recently uploaded (20)

Simulation and 3 d visualization of complex molecular structure for study of protein and nano materials