> I am running unfolding simulations on an 81-redisue protein in explicit
> water (ca. 7500 water molecules). I want to determine the closest 100
> water molecules to the protein during and at the end of the simulation.
> My ptraj script is as follows:
>
> #!/bin/csh
> ptraj file1.prmtop << EOF
> trajin file2.nc
> solvent byres :WAT
> closest 100 :1-81 first
> average near100.pdb pdb
> EOF
>
> The pdb file that is created when I run this script shows only water
> molecules clustered within the nucleotide binding site of the protein.
> There are no water molecules anywhere else around or within the
> molecule. Yet, the starting x-ray structure shows that water molecules
> are distributed evenly around the protein. Is there something wrong with
> my ptraj script? Should I be doing something else?
This is because the average structure of a system with mobile waters will
tend towards zero (the center). You need to look at individual snapshots.
Also, you likely want to make sure the water is imaged as well..
trajin file2.nc
center :1-81 mass origin
image origin center familiar
solvent byres :WAT
closest 100 :1-81 first
Note that the water identity is not maintained... If instead you want to
get an average picture of the solvation, I would use the grid command
(the output of which can be visualized in VMD or Chimera as an X-plor
density map). Alternatively, you can output a trajectory (with the 100
closest waters) and watch a movie of this (for which you'll likely need a
prmtop with 100 waters). FYI on grid input:
trajin file2.nc
center :1-81 mass origin
image origin center familiar
rms first mass out rms.dat :1-81
grid wat.grid 100 0.5 100 0.5 100 0.5 :WAT.O
translate x -0.25 y -0.25 z -0.25
strip :WAT
average avg.pdb pdb
Then, you load up the avg.pdb; without moving it (so the PDB is in the
same reference frame as the grid), load up the grid and then contour to
the desired level. If avg.pdb is not the structure you want to look at,
still load this first, load the pdb of interest and fit to avg.pdb, then
load up the grid...
--tec3
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Aug 10 2010 - 09:30:04 PDT
