VMD-L Mailing List

From: Thomas C. Bishop (bishop_at_tulane.edu)
Date: Thu Apr 24 2008 - 09:40:50 CDT

I break molecs into chains before building structure for molec models.

here's a short VMD script to do that. call the script separate.vmd

then you can separate the pdb into chains A.pdb, B.pdb C.pdb etc...

vmd -pdb rcsbpdbent.pdb -e separate.vmd

(or get fancy and put this line of vmd code in a shell script that takes
a PDBID as an argument.. then "bam" you can download the pdb, separate
the pdb into chains and with a little more scripting build the molecular
model, solvate it, submit it to your local queue for simulation, run
minimization,equilibration, dynamics. conduct standard analysis, e.g.
nucplot, on the simulation and display the results in a web interface...
but I'm getting carried away :-)

###################################################
## TcBishop bishop_at_tulane.edu 4-24-2008.
## this vmd script selects altloc 1
## as the conformation of choice and
### separates the structure into chains

set all [atomselect 0 "altloc 1 '' " ]
set alt1 [atomselect 0 " altloc 1 " ]
set ilist [$alt1 get index ]
foreach i {$ilist } {
  set isel [atomselect 0 "index $i "]
  $isel set altloc { " " }
}

set all [atomselect 0 "altloc ''" ]

set chains [lsort -unique [ $all get chain ] ]

foreach chn $chains {
  set thischain [atomselect 0 "chain $chn and altloc '' " ]
  $thischain writepdb $chn.pdb
}

quit
###########################################################3

On Thu, 2008-04-24 at 16:12 +0530, BERGY wrote:
> dear friends,
> i have 89 protein pdbs bound to DNA. Of which some have more then 3
> protein interacting with DNA and the rest not having contact. i want
> to study the protein DNA interface , so i wish to make files that will
> seperate for each interface. so how to identify the chains that are
> contacting the DNA. Will 5 Angstrom sphereical cut off around each
> Phosphate atom will do it ?? Or should i scan for all possible H2
> bonds and see wat chains are involved? is there a optinon in VMD for
> this to assign the HB cut off for to 5Angstrom and run it for whole
> pdb file.. suggest any other program availabel to study the DNA
> -Protein interface like say NUCPLOT . 

-- 
***********************
*  Thomas C. Bishop   *
*Office: 504-862-3370 *
*CCS:    504-862-8391 *
*Fax:    504-862-8392 *
***********************