Previous Thread
Next Thread
Print Thread
Joined: Mar 2006
Posts: 18
P
Forum Member
OP Offline
Forum Member
P
Joined: Mar 2006
Posts: 18
i will simulate a protein in a box of water. the length of this protein along
one direction is quite long, while lengthes along other directions are much
smaller. so i want to let the protein's long axis coincide with the z-axis
of the box IN simulations so that i can use a box with smaller xy-cross-section,
or fewer water molecules.

it is quite easy to do that off-simulation by 'coor orient', but i wonder if i
can do that IN simultions, and if this will affect the simulation results?
in addition, i wonder if i can let the protein's principle axis coincide with
the box's z-axis IN simulations?

i think one possible way is to constraint some atom (A) in the long axis to
the z axis besides to constraint the center of the protein to the center of
the box. if the protein do not undergo big conformational change in simulations,
the atom A will not deviate from the long axis too much and the long axis will
agree with the z axis roughly. is this idea reasonable? will it affect simulation
results?

thank you so much.

Joined: Sep 2003
Posts: 8,588
Likes: 11
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,588
Likes: 11
Suppressing molecular translation and rotation is usually a bad idea, because it does have an impact. It is quite likely than anything done to keep the protein from moving or rotating or bending will bias the results.

Joined: Mar 2006
Posts: 18
P
Forum Member
OP Offline
Forum Member
P
Joined: Mar 2006
Posts: 18
in my opinion, i do not think suppressing the translation of protein will affect simulation results. i think it is equivalent to put the origin of the reference frame on the COM of protein. suppressing rotation should not affect the results neither in principle, as the protein is put in a box of water, which is supposed to be isotropic. am i correct?

Joined: Sep 2003
Posts: 8,588
Likes: 11
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,588
Likes: 11
Correct in principle, but not in practice. My implication was that you have to be very careful about what steps you take. Fixing some atom coordinates or using positional restraints will certainly change the outcome. Translation and rotation can be rigorously removed, and often are after the fact; that's not the same as fixed atom constraints.

There is a command to keep a protein, etc. centered (see the end of dynamc.doc), but that does not remove rotation. (It also does not support atom selection, and can only be applied to atoms defined first in the PSF.) I don't think there is currently any feature in CHARMM that will periodically re-align the molecule during dynamics. It would certainly make the dynamics somewhat slower to do so, maybe a lot slower.

The system you've described, with a long axis, is highly anisotropic; I don't follow what you mean by the isotropic comment.


Moderated by  BRBrooks, 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: 22 (0.008s) Memory: 0.7419 MB (Peak: 0.7918 MB) Data Comp: Off Server Time: 2022-01-24 04:32:33 UTC
Valid HTML 5 and Valid CSS