An agent-based approach for modeling molecular self-organization
- *Dipartimento di Chimica “G. Ciamician,” Università degli Studi di Bologna, Via F. Selmi 2, 40126 Bologna, Italy; and †Department of Chemistry, Center for Nanofabrication and Molecular Self-Assembly, Northwestern University, Evanston, IL 60208
-
Contributed by Mark A. Ratner, November 15, 2004
Abstract
Agent-based modeling is a technique currently used to simulate complex systems in computer science and social science. Here, we propose its application to the problem of molecular self-assembly. A system is allowed to evolve from a separated to an aggregated state following a combination of stochastic, deterministic, and adaptive rules. We consider the problem of packing rigid shapes on a lattice to verify that this algorithm produces more nearly optimal aggregates with less computational effort than comparable Monte Carlo simulations.
Footnotes
-
↵ ‡ To whom correspondence should be addressed. E-mail: ratner{at}chem.northwestern.edu.
-
Author contributions: A.T., V.W., and M.A.R. designed research; A.T. and V.W. performed research; A.T., V.W., and M.A.R. analyzed data; and A.T., V.W., and M.A.R. wrote the paper.
-
Abbreviation: MC, Monte Carlo.
-
↵ § Alternatively, an agent can be split if its energy does not satisfy the less strict criterion
(i.e., the energy is compared only with the best energy of the agents of the same size).
- Copyright © 2005, The National Academy of Sciences
