Delivered by Ingenta to:
Guest User
IP : 125.46.141.129
Sun, 26 Dec 2010 13:21:31
Copyright © 2011 American Scientific Publishers
All rights reserved
Printed in the United States of America
RESEARCH ARTICLE
Advanced Science Letters
Vol. 4, 74–79, 2011
DNA Self-Assembly for the Minimum
Vertex Cover Problem
Yanfeng Wang
1 2
, Peipei Hu
2
, Xuncai Zhang
1 2
, and Guangzhao Cui
1 2 ∗
1
Research and Development Center of Biological Information Technology,
Zhengzhou University of Light Industry, Zhengzhou 450002, China
2
College of Electrical and Electronic Engineering,
Zhengzhou University of Light Industry, Zhengzhou 450002, China
The problem of finding a minimum vertex cover (MVC) is a classical optimization problem in computer sci-
ence and is a typical example of an NP-complete optimization problem that has attracted a great interest of
researchers because many difficult real-life problems can be formulated as instances of the MVC. The tile
assembly model, a formal model of crystal growth, is of special interest to computer scientists and mathemati-
cians because it is universal, which takes full advantage of the huge parallelism, information is encoded in
DNA tiles and a large number of tiles can be self-assembled via sticky end associations. This paper presents
a method of DNA self-assembly for solving the MVC problem. Several types of DNA tiles are designed for
mapping to the edges of a graph and as the inputs of the tile assembly model, then the system search vertices
based on the nondeterministic choice and get all possible ver tex covers, finally obtain the MVC(s) via gel elec-
trophoresis. The potential computations by DNA self-assembly for the vertex cover problem is promising given
the operational time complexity of On. This work shows fur ther evidence for the ability of DNA self-assembly
to solve NP-complete problems.
Keywords: DNA Computation, Self-Assembly, Minimum Vertex Cover Problem, DNA Tile.
1. INTRODUCTION
In 1994, Adleman successfully solved the Hamiltonian path
problem of seven nodes by using the actual DNA molecules.
1
Following the success, a new research field called DNA com-
putation has been established, and has made a great progress
in the last decades. One of the most significant achievements
of DNA computation has been its contribution to the massive
stream of research in self-assembly, by providing computational
insights into a number of fundamental problems. The research on
self-assembly has some challenges and breakthroughs in recent
years, mainly present in three categories: types of self-assembly
in bulk media, types of components for self-assembly in bulk
media and self-assembly at interfaces.
2
Rothemund and Winfree
have generalized Adleman’s ideas to use an exponential num-
ber of independent nodes, forming a formalized mathematical
model of self-assembly, the tile assembly model.
3
Computation
and self-assembly are connected by the theory of tiling, of which
Wang tiles
4
are the prime example. The tile assembly model is
a model of crystal growth, in which individual components are
square tiles with special labels match. Because that DNA tiles
can be more easily “programmed” to incorporate the constraints
of a given problem, it is possible to exercise some degree of
∗
Author to whom correspondence should be addressed.
controlling, and therefore parallel computation can be enhanced
by self-assembling process, in which information is encoded in
the DNA tiles.
The first experimental demonstrations of computation using
DNA tile assembly were in Ref. [5], which gave a two-layer,
linear assembly of triple-crossover (TX) DNA molecules that
executed a bit-wise cumulative XOR computation. Barish et al.
provided a DNA implementation of a tile system that copies an
input and counts in binary.
6
Cook et al. used the tile assem-
bly model to implement arbitrary circuits.
7
Similarly, Rothemund
et al. demonstrated DNA implementation of a tile system that
computes the XOR function, resulting in a Sierpinski triangle.
8
Brun proposed and theoretically studied the systems that compute
the sums and products of two numbers by using the tile assembly
model.
9
He found that in the tile assembly model, adding and
multiplying can be done using O(1) tiles, and that both computa-
tions can be carried out in a linear time. Then, he combined these
systems to create two new systems with more complex behavior
for solving two NP-complete problems (the satisfiability and sub-
setsum) by using constant computational tile types.
10 11
Inspired
by Brun, we proposed two systems to compute the difference
and quotient of two input numbers using the tile assembly model
with O(1) tiles, and these systems can be carried out in poly-
nomial time. Furthermore, we provided a scheme to factor the
74 Adv. Sci. Lett. Vol. 4, No. 1, 2011 1936-6612/2011/4/074/006 doi:10.1166/asl.2011.1178