Constraint coupling analysis

Compute the largest absolute eigenvalue of the A matrix in the LINCS algorithm for the standard, HMR and VIS CHARMM versions. This repository only contains a short script and example Gromacs files. No installation necessary.

Requirements:

• python3
• numpy
• MDAnalysis

How-to:

1. Depending on your Gromacs and MDAnalysis versions, you might have to re-create the tpr files:

   gmx grompp -f standard.mdp -c standard.gro -p standard.top -o standard.tpr
   gmx grompp -f hmr.mdp -c hmr.gro -p hmr.top -o hmr.tpr
   gmx grompp -f vis.mdp -c vis.gro -p vis.top -o vis.tpr

2. MDAnalysis does not have a separate representation for bonds and constraints. You need to specify a file containing the indices of the bond/constraint to be included in the analysis. You can get a list of constraints from

   • the itp file
   • the tpr file (using gmx dump -s *.tpr)
   • an interactive / debugging MDAnalysis session

   For ease of use, the list of selected constraints is printed to screen. NOTE: the list must be indexed from 0 (see for example constr-all.dat)

3. Run the analysis:

    python constraint-coupling.py datafiles/hmr.tpr datafiles/hmr.gro constr-all.dat

   NOTE: a negative LINCS order ususally means that the eigenvalue is larger than 1.