Previous Thread
Next Thread
Print Thread
Page 2 of 2 1 2
Joined: Sep 2003
Posts: 8,602
Likes: 21
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,602
Likes: 21
If you intend to do a number of different analyses using this approach, it might be a good idea to create new trajectories with the x,y centering already performed. That way you can eliminate the overhead of the image update from subsequent runs.


Rick Venable
computational chemist

Joined: Feb 2009
Posts: 83
B
blubbi Offline OP
Forum Member
OP Offline
Forum Member
B
Joined: Feb 2009
Posts: 83
Originally Posted By: rmv
If you intend to do a number of different analyses using this approach, it might be a good idea to create new trajectories with the x,y centering already performed. That way you can eliminate the overhead of the image update from subsequent runs.


Good point! Thanks a lot!

Cheers
Bjoern

Joined: Feb 2009
Posts: 83
B
blubbi Offline OP
Forum Member
OP Offline
Forum Member
B
Joined: Feb 2009
Posts: 83
Hi again.

I have a question regarding the "update" command.
As long as I am not interested in nonbonded properties like H-Bonds and other electrostatic stuff is it safe to ignore the update command?

Lets say I read the trajectory without setting up the crystal information, iterate it step by step while re-centering every step and then selecting the Lipids under my protein.

Is it necessary to use the "update" command?
And is my assumption right that the update command only updates the nonbonded properties (Which would make no sense when there is no crystal information)

Thanks a lot,
Bjoern

Joined: Sep 2003
Posts: 4,840
Likes: 3
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 4,840
Likes: 3
If you are only doing geometry related analyses I don't see any need for the update command. COOR HBONd also works fine without UPDAte.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Joined: Feb 2009
Posts: 83
B
blubbi Offline OP
Forum Member
OP Offline
Forum Member
B
Joined: Feb 2009
Posts: 83
Thanks for confirming my assumption.

Cheers,
Bjoern

Joined: Sep 2003
Posts: 8,602
Likes: 21
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,602
Likes: 21
That is not correct if you intend to look at protein:lipid interactions, unless you have a new trajectory created with MERGE COOR RECENTER. Regardless of whether the analysis is geometric or energetic, if you move the system to center the protein, if it was protruding from the unit cell, there will be many missing lipids UNLESS you do an image update to repack the lipids. How do you intend to recenter w/o the crystal information?

Also, COOR HBOND is only image aware if if you've set up images.


Rick Venable
computational chemist

Joined: Feb 2009
Posts: 83
B
blubbi Offline OP
Forum Member
OP Offline
Forum Member
B
Joined: Feb 2009
Posts: 83
Thanks a lot for the rectification.

One thing I am curious about:
When running a simulation for a long time it might grow or shrink over time so the crystal data (side length of box in particular) I used to setup the simulation is "outdated".

When using "correl" or "update" with crystal information I ran into the following error:
Code:
      ***** LEVEL -1 WARNING FROM <TSTCRD> *****
      ***** SOME ATOM COORDINATES UNDEFINED OR OUT OF RANGE
      ******************************************
      BOMLEV (  0) IS REACHED - TERMINATING. WRNLEV IS  5

Here's the example code:
Code:
! Read in Topology and  Parameter files
stream ../read_psf_and_toppar.str

! Setup Crystal
crystal define TETR 93.7831 93.7831 144.964 90 90 90
open unit 1 read card name crystal.xtl
crystal read unit 1 card
close unit 1

! Image centering
image byresid XCEN 0.0 YCEN 0.0 ZCEN 0.0 sele ALL .and. .not. segid PRO* .and. .not. segid RNAA end
image byseg XCEN 0.0 YCEN 0.0 ZCEN 0.0 sele segid RNAA .or. segid PRO* end

open unit 99 file read name concatenated_recentered.dcd
traj query unit 99

calc NSTEP ?nstep
set START ?start
calc STOP ?nstep
set SKIP ?skip

calc mxa 9164 + 40334
calc mxs 6
calc mxt @NSTEP

correl maxtim @MXT maxa @MXA maxs @MXS

define MS sele segid MEMB end
define PS sele segid PRO* end

end
stop


The crystal.xtl was written with the "write crystal" command after solvating, neutralizing, centering and shaping the system.

When I omit the "read crystal.xtl" line, the error is gone.

Adding the "noupdate" option to "correl" omits the error as expected.

Or does this not at all relate to the side length of the box?

Cheers,
Bjoern

Last edited by blubbi; 01/19/11 05:27 PM.
Joined: Sep 2003
Posts: 4,840
Likes: 3
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 4,840
Likes: 3
This is completely unrelated to your box and your trajectory. correl can be used for energy related analyses, and by default calls the update routine to have various lists filled. Since in your case there are no coordinates present the error message you see is issued. You correctly added the noupdate keyword to avoid this; you could also have read in coordinates prior to entering the correl module, but the noupdate method is faster and cleaner.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Joined: Feb 2009
Posts: 83
B
blubbi Offline OP
Forum Member
OP Offline
Forum Member
B
Joined: Feb 2009
Posts: 83
While checking my other scripts I realized my mistake.

I simply missed to read in the initial coordinates as I do usually.

Your answer came to fast.
Sorry for the dumb mistake and bothering the forum with it.

But it's good to know that adding the "noupdate" option is the more appropriate way.

Thanks a lot - again -
Bjoern

Page 2 of 2 1 2

Moderated by  lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5
(Release build 20201027)
Responsive Width:

PHP: 7.3.31-1~deb10u1 Page Time: 0.010s Queries: 33 (0.005s) Memory: 0.7791 MB (Peak: 0.8696 MB) Data Comp: Off Server Time: 2022-08-16 10:41:13 UTC
Valid HTML 5 and Valid CSS