Polybuild, corruption of 'polymer builder', is a software developed in Shell Script language, interpreted by GNU bash (version 4.3.30), which facilitates building initial configurations for molecular dynamics simulations of dispersed polymers.
Based on the concise specification of the microstructure of a linear architecture dispersed polymer system, the quantities of different molecules in the simulated sample and the parameters related to the simulation box, Polybuild creates the configuration files, called xyz files, containing the positions of all particles for all molecular types in the sample. After that, it creates a file in a format recognized by the software Playmol[1], which packs the molecules in a simulation box with all the required description.
Particle-based simulations require an initial configuration, which consists in the initial positions of all particles of the simulated system. The task of creating such initial conditions may not be trivial. It consists of an optimization problem with several constraints, such as: (i) box size and shape restrictions, (ii) reproduction of the system density, and (iii) the most difficult one, as short-range repulsions should not br excessively large, to prevent in numerical instability during the course of the simulation.[2,3] It is a common practice to use computer programs that solve this optimization problem by packing the particles in simulation boxes that are subject to the constraints that force fields impose.[4] Generally, these programs create an initial configuration by replicating atoms and molecules that make up the system in a defined region of space, respecting a series of spatial constraints, thus creating complex initial configurations. However, the number of different species in a dispersed polymer system can be in the order of many thousands. This requires, therefore, the description of the topology of these many thousands of chains. Even if the simulation reproduces only a small portion of space or if it involves a coarse-grained models, obtaining such information and feeding it to computer programs that generate initial configurations can be costly.
Here a coding system is proposed for linear chains. Each block is represented by two pieces of information: the first one specifies the type of particle that constitutes the block and the second represents the block length, which must necessarily be a cardinal number. For example, 'A 5 B 5' represents a diblock-type oligomer, consisting of a block with five beads of type A attached to a block with five beads of type B. This coding is a representation of the structural formula of the molecule, so that it encloses all compositional information and the connectivity between the component particles of the molecule. However, it does not carry any conformational information, which is also necessary for the specification of the molecule to be replicated. Therefore, in addition to the sequence that translates the structural information, it is necessary to specify conformational parameters for writing the xyz files. The simplest choice is adopted, assuming that insertes particles form a straight line in a given cartesian direction. Thus, each particle is inserted at a distance lb from the previous one, forming a connecting angle of 180º.
- Playmol - see reference [1].
- a bash interpreter (the version provided here is tested for a GNU bash version 4.3.30).
The use is straightforward. The 'polybuild.sh' script receives an input file that contains the microstructural specifications, system parameters, that must and Playmol parameters to be used for packing - see the template file called 'input_file_template'.
The command that processes the script has the following syntax and is capable of receiving multiple input files at once:
> ./polybuild.sh input_file1 [input_file2 … input_fileN]
If this script is used in the preparation of scientific publications, the following paper should be cited.
LEMOS, T. S. M., ABREU, C. R. A., PINTO, J. C. Mesoscopic Simulation of Dispersed Copolymers - Effects of Chain Length, Chemical Composition, and Block Length Distributions on Self-Assembly, Macromolecular Theory and Simulations, 2019
[1] C. R. Abreu, "Playmol", http://atoms.peq.coppe.ufrj.br/playmol.
[2] D. Frenkel, B. Smit, Understanding Molecular Simulation, Academic Press, London, 2001.
[3] M. P. Allen, D. J. Tildesley, Computer Simulation of Liquids, Oxford University Press, New York, 1991.
[4] L. Martínez, R. Andrade, E. G. Birgin, J. M. Martínez, Journal of Computational Chemistry 2009, 30, 2157-2164.