GridMAT-MD:Algorithm

The algorithm of the GridMAT-MD program is discussed in greater detail in the corresponding publication. Here, we will give an overview of the main points of the calculations.
  • Bilayer Thickness calculation:
    1. Based on a user-defined reference atom, each lipid in the top leaflet is assigned a paired lipid in the bottom leaflet, based on proximity in the x- and y- directions. The z-distance between these points is calculated. Periodicity is taken into account.
    2. The same step is repeated, only using the bottom leaflet as the reference, measuring the z-distance to the top leaflet.
    3. The distances with top and bottom leaflets as references are averaged, and the final output written to a generic ASCII .dat file.

  • Area Per Lipid Headgroup calculation:
    1. The same calculations as in the thickness calculation are performed in order to assign thickness values to each grid point. A greater number of grid points is generally recommended to achieve good resolution.
    2. Protein atoms found within the lipid headgroups are assigned to grid points; this total area will be subtracted from the lateral area of the bilayer.
    3. The size of the system is determined from the dimensions of the unit cell.
    4. The recurrence of a thickness value for a given grid point determines the fraction of the total grid occupied by a single lipid. Based on this fraction, the area is calculated as that fraction of the total lateral area of the unit cell.
    5. The area for each individual lipid is written to a generic ASCII .dat file.
The final data output for the thickness calculation is written as a raw data file, giving a single column list of thickness measurements. We have successfully interfaced the output of this program with XMatrix, a component of the Matpack software package, as well as Gnuplot. For data usage, please refer to the documentation provided with the program.

Area per lipid data is also printed to a generic .dat file, with the area of each lipid residue printed alongside the corresponding molecule. This format allows for simple, manual statistical analysis, or processing with standard *nix utilities like awk or Perl.

GridMAT-MD Home
Bevan Lab Homepage

Site design copyright 2008-2011 by Justin Lemkul
Questions? Problems with the site? Comments? Send them to the Webmaster