Previous Thread
Next Thread
Print Thread
Joined: Jul 2009
Posts: 6
C
Forum Member
OP Offline
Forum Member
C
Joined: Jul 2009
Posts: 6
Hi guys,

I was wondering how to calculate the optimal number of water molecules for an MD simulation of a protein. Apparently there is no risk of adding too many water molecules because at the saturation point the excess waters overlap with each other and the protein and must be deleted. But is there another reason why I should avoid of adding too many waters?

Last edited by Conjugated; 10/07/09 11:58 AM.
Joined: Sep 2003
Posts: 4,863
Likes: 10
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,863
Likes: 10
Cost.

If you use a compact primary cell (eg RHDO) for your PBC and make it 10-15A larger than the solute on each side, you effectively have a solute at infinite dilution - the solute would then never be closer to an image of itself than 20-30A (unless there are large conformational changes). Well before 20A there is no structure left in the radial distribution of water oxygens.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Joined: Dec 2005
Posts: 1,535
Forum Member
Offline
Forum Member
Joined: Dec 2005
Posts: 1,535
I've been pondering for a while on a related question. In our lab, we always solvate globular proteins in a truncated octahedron (OCTA). Theoretically spoken, a rhombic dodecahedron (RHDO) should offer a slightly better packing factor (ie. less water needed for the same distance between the solute and its nearest images). Are there any practical disadvantages to using RHDO?

All I got so far is a rumor that there's some computational inefficiency associated with it...

Joined: Sep 2003
Posts: 4,863
Likes: 10
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,863
Likes: 10
No - RHDO is simpler, and nothing demonstrated against. I personally get nervous when one of the parameters must be given with a zillion decimals. Rumours - now, haven't we been trained at expensive universities for many years to have a specific attitude towards the rumour-reality dichotomy...
The small difference in packing density is not significant.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Joined: Dec 2005
Posts: 1,535
Forum Member
Offline
Forum Member
Joined: Dec 2005
Posts: 1,535
Thanks Lennart, that's just what I needed to know!

Joined: Jul 2009
Posts: 6
C
Forum Member
OP Offline
Forum Member
C
Joined: Jul 2009
Posts: 6
Okay, thanks for the PBC advise. But the question I was wondering was the water molecule concentration within the box. I mean, should I just generate a PDB file with random coordinates for minimum number of TIP3 molecules, and then use CHARMM to strip out the overlapping waters? And how can I predict the minimum number of water molecules, given that I know the dimensions of the protein?

And in general, how do you usually prepare the explicit water? Is there an easier way than to read a randomly generated PDB file, like: generate 9999 TIP3P molecules and setup their coordinates randomly in CHARMM?

Joined: Sep 2003
Posts: 4,863
Likes: 10
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,863
Likes: 10
Ther are scripts to do this in the Script Archive.
To get the correct density you can run a constant pressure MD simulation after the initial setup.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden

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~deb10u2 Page Time: 0.011s Queries: 28 (0.007s) Memory: 0.7578 MB (Peak: 0.8201 MB) Data Comp: Off Server Time: 2023-01-27 07:19:22 UTC
Valid HTML 5 and Valid CSS