International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 03 Issue: 03 | Mar-2016 www.irjet.net p-ISSN: 2395-0072
© 2016, IRJET | Impact Factor value: 4.45 | ISO 9001:2008 Certified Journal | Page 671
Metaheuristic Techniques for Conformational Search
SIEW MOOI LIM
1
, MD. NASIR SULAIMAN
2
, NORWATI MUSTAPHA
3
, ABU BAKAR MD. SULTAN
4
FACULTY OF COMPUTER SCIENCE AND INFORMATION TECHNOLOGY, UNIVERSITI PUTRA MALAYSIA, MALAYSIA
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract The drawback in conformational search (CS) is in
locating the most stable conformation of a molecule with
the minimum potential energy based on a mathematical
function. The number of local minima grows exponentially
with molecular size and this makes it that more difficult to
arrive at a solution. It had been confirmed that CS belongs
to the category of NP-hard (non-deterministic polynomial
time) problem. Such complexity requires an equally long
amount of time to achieve resolution. This phenomenon is
thus known as the 'combinatorial explosion'.
Metaheuristic techniques have been constantly used in
solving CS problems. These population-based probabilistic
techniques explore conformational space by random
perturbation of atomic Cartesian coordinates or the torsion
angles of rotatable bonds. These methods focus on
exploring a search space with maximum efficacy. With one
or more solutions in the beginning, metaheuristic method
follows with a more iterative approach to optimize the
search in promising areas away from local solutions. This
method is often employed in circumstances where the exact
solution methods are unfeasible within a limited time frame.
As such, this paper presents various past metaheuristics
approaches that have been brought forth in regards to the
problem of an effective exploration of the conformational
states of molecular systems. Each metaheuristic method is
accompanied by its advantages and disadvantages. The
concepts of each approach will be explained and their
respective applications are discussed.
Key Words: Model building methods, distance geometry,
smoothing methods, systematic search
1. INTRODUCTION
Conformational search (CS) is a term familiar to those in
the field of applied mathematics and computational
chemistry. CS is mathematically represented as a
continuous global optimization problem. In CS, the
variables are the torsion angles or coordinates that are
used to represent the conformation of the molecule (e.g.
polypeptide chain). The objective function value is the
potential energy function. By varying the values of the
variables, the global minimum value of the objective
function can be achieved; that is to locate the most stable
conformation of a molecule with the minimum potential
energy. Years of research have seen CS performing
important and widely used molecular modeling
applications which include flexible rings and macrocycles
molecules [1], cyclic, acyclic, mixed single molecules and
host-guest complexes [2], molecules under inhibited
conditions, peptide and protein folding [3], simulations of
protein-ligand docking [4, 5] and various drug design
applications such as Quantitative Structure Activity
Relationship (QSAR), virtual screening of virtual libraries
and active analog approaches [6].
The exact methods are unable to solve the complexity of
CS problems. Therefore, metaheuristic techniques [7]
including genetic algorithm have become mandatory and
has attracted considerable attention especially from
evolutionary algorithm community.
We have introduced a novel real coded genetic algorithm
which is capable in solving two CS application problems i.e
minimizing a molecular potential energy function and
finding the most stable conformation of pseudoethane
through a molecular model that involves a realistic energy
function [8-12]. The focus of this paper has been pivoted
on the five major categories of approaches including five
common metaheuristic techniques that have been applied
to CS problems.
2. Application of Metaheuristic
Techniques to Conformational Search
Problems
Metaheuristic techniques include all stochastic algorithms
with randomization and local search. Randomization
allows a shift from the algorithms in a local search to the
global scale. Therefore, almost all metaheuristic
algorithms were designed to suit the global optimization.
Metaheuristic algorithms consist of two major
components namely exploration (diversification) and
exploitation (intensification). GA, Tabu search and
particle swarm optimization are reported to heavily study
on maintaining a good balance between exploration and
exploitation in preserving diversity [13, 14].
The function of exploration is such that it generates
diverse solutions in order to explore the search space on
the global scale. On the other hand, exploitation draws
upon the local search region, in which the information of a
current good solution in this region will be further