Getting Dihedral Angles From Simulations

MutInf analyzes molecular dynamics simulations by using internal (BAT) coordinates, as opposed to Cartesian coordinates, based on the intuition that backbone and sidechain torsion angle motions drive biologically relevant fluctuations and conformational changes. I recommend running at least 6-10 simulations (different random number seeds) of 10+ns each with snapshots written at 1ps intervals. Alternatively, six “blocks” of 50ns each could be taken from three 100ns simulations. If the structure used is not from a crystal structure (even if there is one mutation), I recommend not using the first 5ns or so, as this will likely reflect unequilibrated data. See option “-z <step number>” to skip steps during analysis (if they weren’t skipped over in post-processing).

You will need dihedral angle files for each torsion, containing two space-delimited fields:

1) time or timestep or any number, and

2) dihedral angle in degrees.

I use the GROMACS utility “g_chi” to produce these files.

These files can be named .xvg or .xvg.gz, for each torsion for each residue. for example:

chi1PHE42.xvg.gz

chi2PHE42.xvg.gz

phiPHE42.xvg.gz

psiPHE42.xvg.gz