Re: how to fix the problem

From: Jeff Comer (jeffcomer_at_gmail.com)
Date: Sat Nov 15 2014 - 08:33:54 CST

You can select the broken part of your protein by using a geometric
selection. Something like "protein and z>50". It will depend on your
system geometry.

Assuming that your basis is orthonormal (your system is a rectangular
prism), you can get the basis vectors at each frame of the dcd using a
script like the following:

set sel [atomselect top "protein"]
set frames [molinfo top get numframes]
for {set f 0} {$f < $frames} {incr f} {
  molinfo top set frame $f
  set lx [list [molinfo top get a] 0 0]
  set ly [list 0 [molinfo top get b] 0]
  set lz [list 0 0 [molinfo top get c]]

  # Shift how you want.
  # This is just an example.
  if {[lindex [$sel get z] 0] >50} {
     $sel moveby [vecinvert $lz]
  }
}
# Write the resulting dcd.
set all [atomselect top all]
animate write dcd out.dcd beg 0 end -1 sel $all top

–––––––––––––––––––––––––––––––––––———————
Jeffrey Comer, PhD
Assistant Professor
Institute of Computational Comparative Medicine
Nanotechnology Innovation Center of Kansas State
Kansas State University
Office: P-213 Mosier Hall
Phone: 785-532-6311

On Fri, Nov 14, 2014 at 10:07 PM, Liqun Zhang <lxz79_at_case.edu> wrote:
> Hi, Jeff:
>
> Thank you very much for your information. In my simulation, the box size
> could fluctuate. So how to decide how much to shift, and also how to select
> the broken section of the protein? Thanks a lot.
>
> LQZ
>
> On Fri, Nov 14, 2014 at 9:02 PM, Jeff Comer <jeffcomer_at_gmail.com> wrote:
>>
>> Oops, I meant "$sel moveby $vector".
>>
>> –––––––––––––––––––––––––––––––––––———————
>> Jeffrey Comer, PhD
>> Assistant Professor
>> Institute of Computational Comparative Medicine
>> Nanotechnology Innovation Center of Kansas State
>> Kansas State University
>> Office: P-213 Mosier Hall
>> Phone: 785-532-6311
>>
>>
>> On Fri, Nov 14, 2014 at 8:02 PM, Jeff Comer <jeffcomer_at_gmail.com> wrote:
>> > Hi LQZ,
>> >
>> > If you don't like where the pbc join put the protein, just shift it
>> > where you would like it by a cell basis vector (using $sel move
>> > $vector). Shifting by a cell basis vector or any linear combination of
>> > cell basis vectors with integer coefficients doesn't change the system
>> > physically.
>> >
>> > Forgive me if you already understand this, but I should point out that
>> > various pbctools commands only change how the system looks and don't
>> > change it in any physically meaningful way. In general, how you wrap
>> > your structure doesn't matter to NAMD. For example, even if VMD makes
>> > it look like a bond crosses almost the entire system except for 2
>> > angstroms, NAMD considers only the minimum distance between two
>> > periodic images, which is just 2 angstroms apart.
>> >
>> >
>> > –––––––––––––––––––––––––––––––––––———————
>> > Jeffrey Comer, PhD
>> > Assistant Professor
>> > Institute of Computational Comparative Medicine
>> > Nanotechnology Innovation Center of Kansas State
>> > Kansas State University
>> > Office: P-213 Mosier Hall
>> > Phone: 785-532-6311
>> >
>> >
>> > On Fri, Nov 14, 2014 at 5:19 PM, Liqun Zhang <lxz79_at_case.edu> wrote:
>> >> Dear all:
>> >>
>> >> I have run a simulation for a protein in lipid system for a very long
>> >> time.
>> >> In order to remove the pbc effect, I unwrapped the trajectory for
>> >> protein
>> >> and the lipids using pbc commands in vmd program, which made part of
>> >> the
>> >> protein broken in the z-direction. But when I tried to use join command
>> >> to
>> >> join the broken part to the major part of the protein, the major part
>> >> of the
>> >> protein instead was shifted to the broken part. I am not sure how to
>> >> solve
>> >> the problem. Can I get some suggestion here? Thank you very much.
>> >>
>> >> Best,
>> >>
>> >> LQZ
>
>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:22 CST