Computational Modelling Group

Dual resolution simulations of lipid membrane systems

Research Team
Kieran Selvon
Jonathan Essex

A snapshot of a 128 ELBA lipid (DOPC) membrane system, ran in the LAMMPS software.

To enter into the main circulation of the body, all medicine molecules must ultimately cross the lipid bilayer which surrounds the cells in our body. This ability to permeate the lipid bilayer is difficult and costly to asses experimentally. In this project lipid membrane systems are simulated using molecular dynamics with the goal of perfecting a method to elucidate the mechanisms of permeation, as well as quantifying the ability of molecules to pass across the bilayer.

The simulations are performed using the dual resolution ELBA model for lipid and water membrane systems. In this model the water and lipid molecules in the system are treated as coarse grain beads, whilst the drug molecules in the system are treated as atoms. In principle resulting in simulations with accuracy similar to all atom methods, and efficiency similar to coarse grain approaches.

The simulations also make use of 'enhanced sampling' techniques, which speed up otherwise slow events in the simulation, in particular a method called 'Umbrella sampling'. An advantage of this method is that it allows one long simulation to be effectively split into many shorter simulations, all of which can be ran in parallel, significantly improving computational efficiency. The simulations are being ran on the southampton IRIDIS cluster utilising 31 nodes/496 processors at a time, a feat which would not be possible without a super computer.


Life sciences simulation: Biomolecular simulations

Algorithms and computational methods: Molecular Dynamics

Programming languages and libraries: C++, Python

Computational platforms: Iridis, Linux

Transdisciplinary tags: Complex Systems