Biophysics .co.uk

What is LAMMPS?

admin — Wed, 21 Mar 2012 14:33:09 +0000

There are plenty of Molecular Dynamics packages out there, but what is LAMMPS and why could or should you use it and how does it differ from say Amber or CHARMM? Here are some of the LAMMPS features that I have lifted from the front page of LAMMPS at http://lammps.sandia.gov/:

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator.
LAMMPS has potentials for soft materials (biomolecules, polymers) and solid-state materials (metals, semiconductors) and coarse-grained or mesoscopic systems. It can be used to model atoms or, more generically, as a parallel particle simulator at the atomic, meso, or continuum scale.
LAMMPS runs on single processors or in parallel using message-passing techniques and a spatial-decomposition of the simulation domain. The code is designed to be easy to modify or extend with new functionality.
LAMMPS is distributed as an open source code under the terms of the GPL. The current version can be downloaded here. Links are also included to older F90/F77 versions. Periodic releases are also available on SourceForge.
LAMMPS is well documented and supported by Sandia and flexible enough for most purposes.
Importantly for some fields, it includes rigid body rotation which enables orientation properties of atoms and molecules to be included.

You can download the LAMMPS package (FREE from here): http://lammps.sandia.gov/download.html

Calculating mean squared displacement for membranes

admin — Wed, 21 Mar 2012 14:25:54 +0000

Often one wants to understand how best to determine the phase of a simulated lipid membrane. One way is to use the Mean Squared Displacement (MSD) method where one measures how far lipids move with respect to a centre of mass. This way its easy to determine whether your membrane is in a gel or liquid phase. Gel phase lipids would expect to have little movement and no net displacement, however fluid membrane lipids are able to move and hence there is an increase in displacement with time.

We have included the MSD code that works with LAMMPS output from: https://github.com/biomembranes/SmartAnalyzer which contains the python code msd.py which performs this temporal analysis. You’ll want to use it in conjunction with cfg.py where you will want to have details of your trajectories generated with LAMMPS. You can view the readme files here: https://github.com/biomembranes/SmartAnalyzer/blob/master/README

How to find clusters from your molecular dynamics simulations

admin — Tue, 20 Mar 2012 18:00:18 +0000

Given a series of molecules or atoms – how do we define clusters? Well there are a number of approaches that one can look at from all kinds of fields. One could even use machine learning to recognize clusters, however being a pragmatist, one finds the simple method is often the best, rather than spending time coming up with elaborate solutions. Hence one goes to a proximity based algorithm based on nearest neighbor separation.

Aggregation properties are calculated utilizing a recursive clustering algorithm that begins with an arbitrary molecule and subsequently determines its nearest neighbors that are deemed to be on the same cluster. Each one of these nearest neighbors is then used as the starting point for the next nearest neighbor search until the recursive tree is exhausted and no more molecules are found. Once the molecules in an aggregate have been exhausted, a linear search finds a molecule that is not part of any pre-existing aggregate and the procedure repeats. The nearest neighbor search only includes molecules within a radius limited to close neighbors. The algorithm is robust against a wide range of nearest neighbor cut-offs. But you must use this at your own risk. Here are some things you will and must play/experiment with:

The clustering target molecule/atom – used for comparison of the separation (i.e. choose the atom on the molecule that is of interest). Typically one can use the midpoint or the centre of mass of the molecule or whatever seems most appropriate. You can adjust this as desired.
The cut-off radius which is the limit under which molecules/atoms are included (how far apart another atom/molecule can be).

You can find out more from here: http://www.biophysics.co.uk/lammps-tools-for-membranes/clustering-analyser-for-lammps/ and download code based around LAMMPS file formats from here: https://github.com/biomembranes/Cluster – You could adapt the code to almost any file format that you desire. Also be careful regarding the stack depth for many recursive calls.

Computing the dipole orientation of a cluster in VMD

admin — Mon, 19 Mar 2012 18:45:46 +0000

One tool that has generic use beyond just membranes is the dipole viewer code for VMD. The TCL/TK script aims to analyse a cluster over a time range and determine the dipole orientation and properties such as orientation with respect to a lipid tail. You might find some of the code excerpts useful for understanding how to compute and histogram dipole orientations in VMD.

You can find the code located at our GitHub location or select from the following link: https://github.com/biomembranes/VMDDipoleViewer

Building a lipid membrane from scratch

admin — Wed, 29 Feb 2012 16:05:25 +0000

We wanted to be able to build multiple component lipid membrane systems for simulation with the Molecular Dynamic Framework called LAMMPS, so we set about building a series of routines cut in Python that would help us and anyone build systems – solvated or otherwise with relative ease. Of course, you will need to have an understanding of the LAMMPS framework, but there are some example systems included that will let you use the coarse grained lipid models that we currently use.