[gmx-users] contacts per residue in g_mindist

Justin Lemkul jalemkul at vt.edu
Thu Nov 13 19:38:25 CET 2014

On 11/13/14 10:04 AM, serena leone wrote:
> Dear GMX Users,
> I have been trying to analyze the evolution of hydrophobic contacts
> during an unfolding simulation. To do this, I have prepared an index
> file with the hydrophobic residues in the two portions of the protein
> that I simulated and I issued g_mindist as follows:
> g_mindist -f prot.xtc -s prot.tpr -d 0.56 -o atm-pair_hydrophobic -on
> numcont_hydrophobic -or mindistres_hydrophobic.xvg -n protein.ndx -dt 10
> -printresname -respertime -od mindist_hydrophobic.xvg
> If I understood correctly, atm-pair_hydrophobic gives me the atom
> couples that are closest per frame according to their numbering in the
> topology, whereas mindistres_hydrophobic.xvg with -respertime gives me
> the evolution of the minimum distance between each residue in group 1 of
> my analysis with the COM of group 2 of my analysis. Now the two
> questions:
> 1) the mindistres_hydrophobic.xvg file that i open with grace has
> several data sets in it, but the legend looks something like:
> [...]
> @ legend length 2
> @ s0 legend "helix_hydrophobic-34-90_hydrophobic"
> # PRO9 PHE10 LEU14 GLY15 PHE17 ALA18 VAL19 ILE25
> [...]
> The numbering of the residues has been (i believe) taken from the
> topology when I made the index, but this numbering is wrong, the
> numbering in the structure file is exactly n+1 (i.e. PRO10 PHE11
> LEU15...). If I visually analyze the corresponding .gro file, the
> numbering is correct, coherent with the starting pdb, and most important
> it is correct in every other ndx generated for instance by g_bonds,
> index making and so on. Does anybody have any idea on why they appear
> shifted in this case?

Probably because arrays in the code are zero-indexed.

> 2)is there any way to extract the distance of the residue in group 1 not
> from the COM of group 2, but from individual residues in it? in other
> words, a sort of hbonds-style contact map with an index?

g_dist can do this.


> As a minor issue, I assume that PRO9 etc. are the data series names, but
> they do not automatically appear on the plot (in grace). Although I can
> fix this by hand, it might become pretty tedious if the groups become
> bigger. Any hint on how to solve it?
> Thanks everybody and have a great day,
> Serena


Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441


More information about the gromacs.org_gmx-users mailing list