[gmx-users] trjorder, what and how does es count (amount solvent atoms or molecules)?

Justin A. Lemkul jalemkul at vt.edu
Sat Mar 12 17:00:47 CET 2011

Henri Mone wrote:
> Dear Gromacs Users,
> I'm using "trjorder" with the following options:
> $ trjorder -f systemWithSolv.xtc -n system.atom.ndx -s system.pdb -r
> 0.35  -nshell system0.35.nshell.xvg
> As the first group I select all the atoms of a single TYR residue,
> which is part of a larger protein. As the second group I select all
> the solvent atoms (TIP3P). For a radius of 3.5 A I observe in most
> frames the number 8, the observations range from 0 to even 20. What
> does this number mean exactly?
> For TIP3P the OH2-H1 and OH2-H2 distance is approx. 1A and the H1 and
> H2 distance is approx. 1.6A.  This gives a "volume" of 4.2A^3 for
> TIP3P. A sphere of 3.5A would have a volume of 14.6A^3. As a first
> order approximation this would suggest for non buried TYR residue
> about 3 to 4 TIP3P molecules which could fit in this sphere. This
> differs from the number 8 which trjorder measures.
> My questions are:
> 1. What does trajorder count? Does it count the number of solvent
> atoms or of solvent molecules?

trjorder -h:

"With option -nshell the number of molecules within a shell of radius -r
around the reference group are printed."

> 2. How does trajorder count? If trajorder counts the amount of
> molecules. Is it sufficient that a single atom of the molecules is in
> the sphere to be included in the counting or must all atoms of the
> molecule be in the sphere to be included in the counting?

Check the code.  If nothing else, your approximation is off.  Would a sphere of 
0.35 nm completely encompass your whole Tyr residue?  The distance are 
presumably based on a radius of 0.35 nm of each of the constituent atoms, making 
for a somewhat larger effective volume.


> Thanks,
> Henri


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list