Re: psf generation for multiple trehalose molecules

From: Ritu Arora (ritu.arora_at_mail.concordia.ca)
Date: Mon Nov 09 2020 - 04:12:07 CST

Hi Peter,

Omitting 'resetpsf' inside the loop helped. 'tmp.pdb' file looks fine, but all the molecules get stacked on top of each other in 'new.pdb' file.

Ritu
________________________________
From: Peter Freddolino <petefred_at_umich.edu>
Sent: Monday, November 9, 2020 1:08 AM
To: <namd-l_at_ks.uiuc.edu> <namd-l_at_ks.uiuc.edu>; Ritu Arora <ritu.arora_at_mail.concordia.ca>
Cc: Josh Vermaas <joshua.vermaas_at_gmail.com>; namd_at_ks.uiuc.edu <namd_at_ks.uiuc.edu>
Subject: Re: namd-l: psf generation for multiple trehalose molecules

I suspect that the 'resetpsf' inside your loop isn't helping.
Best,
Peter

On Sun, Nov 8, 2020 at 10:53 PM Ritu Arora <ritu.arora_at_mail.concordia.ca<mailto:ritu.arora_at_mail.concordia.ca>> wrote:
Hi Josh,

Thank you for clarifying this. After removing alias commands and minor typos, the following script saves the coordinates and structure of only the last trehalose molecule in the output files. Something is still missing!

package require psfgen
topology top_all36_carb.rtf
mol new carb_all.pdb
set sel [atomselect top "all"]
for { set i 0 } { $i < 150 } { incr i } {
resetpsf
segment L$i {
residue 1 AGLC
residue 2 AGLC
}
patch 11aa L$i:1 L$i:2
$sel set segname L$i
$sel moveby [list 1 1 1]
$sel writepdb tmp.pdb
regenerate angles dihedrals
coordpdb tmp.pdb L$i
}
guesscoord
writepdb new.pdb
writepsf new.psf

Sincerely,
Ritu
________________________________
From: Josh Vermaas <joshua.vermaas_at_gmail.com<mailto:joshua.vermaas_at_gmail.com>>
Sent: Sunday, November 8, 2020 9:13 PM
To: NAMD list <namd-l_at_ks.uiuc.edu<mailto:namd-l_at_ks.uiuc.edu>>; Ritu Arora <ritu.arora_at_mail.concordia.ca<mailto:ritu.arora_at_mail.concordia.ca>>
Cc: namd_at_ks.uiuc.edu<mailto:namd_at_ks.uiuc.edu> <namd_at_ks.uiuc.edu<mailto:namd_at_ks.uiuc.edu>>
Subject: Re: namd-l: psf generation for multiple trehalose molecules

Hi Ritu,

Why are you aliasing at all? In patches that connect more than one residue, the first digit indicates to psfgen which residue the atom name is on. This is fundamentally why atom names can't start with a digit. I think your script otherwise looks close to what I would write.

Josh

On Sun, Nov 8, 2020, 5:26 PM Ritu Arora <ritu.arora_at_mail.concordia.ca<mailto:ritu.arora_at_mail.concordia.ca>> wrote:
Dear All,

I wish to run a simulation of 150 molecules of trehalose. I adapted a psfgen script from earlier NAMD threads, but I am not able to get the script working. The trial script for the carbohydrate section of my system follows:

package require psfgen
topology top_all36_carb.rtf
mol new carb_only.pdb
set sel [atomselect top 'all']
for {set i 0} {$i<150} {incr i} {
resetpsf
segment L$i {
residue 1 AGLC
residue 2 AGLC
}
pdbalias atom 'AGLC 1' O1 1O1
pdbalias atom 'AGLC 1' C1 1C1
pdbalias atom 'AGLC 2' C1 2C1
#
patch 11aa L$i:1 L$i:2
$sel set segname L$i
$set moveby [list 1 1 1]
$set writepdb tmp.pdb
regenerate angles dihederals
coordpdb L$i tmp.pdb
}
guesscoord
writepdb $name.pdb
writepsf $name.psf

The confusion is that there are two AGLCs in trehalose and each has either resid 1 or 2, rather than sequential increments for whole molecule of trehalose. The pdbalias command also seems to be wrong.

I also tried to split the two fragments of trehalose in two separate pdbs:

pdbalias atom 'AGLC 1' O1 1O1
pdbalias atom 'AGLC 2' C1 2C1

for {set i 0} {$i < 150} {incr i} {
segment TRE1$i {pdb output/carb1.pdb }

segment TRE2$i {pdb output/carb2.pdb }

patch 11aa TRE1$i:1 TRE2$i:2
coordpdb output/carb1.pdb TRE1$i
coordpdb output/carb2.pdb TRE2$i
}
I would be thankful for any suggestions to improvise the script.

Best,
Ritu

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:14 CST