[gmx-users] virtual sites and heavy atoms for longer time steps

[Hassan Shallal]

I have some issues to discuss with the mailing list users,
 
1- Is using virtual sites alone enough to allow me to lengthen the time step from 2 fs to 4 fs, or I have also to use heavy hydrogens when creating topology files in order to use 4 fs time step?
 
2- Another question, does vitual sites (assigning virtual hydrogens a mass of zero and adding their masses to the connected heavy atoms) opposes the effect of heavy atoms (increases the mass of the hydroen atoms) or vice versa. 
Or is the situation that heavy atoms treatment applies  when virtual sites treatment is not implemented, and we end up with a systme where some hydrogen atoms have no masses at all (those as virtual sites) whereas other hydrogen atoms are much heavier (those treated as heavy hydrogens)?
 
3- Implementing the scenario of some hydrogen atoms having zero masses and others having heavier masses, how would that affect the accuracy of predicting some energetic aspects like H-bonding? I know there is always a trade between efficiency or fastness and the accuracy because approximations usually fasten the job while decreasing the accuracy, but from experienced users point of view, how dangerous that approximation would be?
 
Any answers or clarification of the above issues will be so much appreciated
 
Hassan

________________________________

From: gmx-users-bounces at gromacs.org on behalf of Roland Schulz
Sent: Wed 12/1/2010 2:15 PM
To: Discussion list for GROMACS users
Subject: Re: [gmx-users] Cut-offs using CHARMM27 ff




On Wed, Dec 1, 2010 at 4:30 PM, Hassan Shallal <hshallal at pacific.edu> wrote:


	Dear Gromacs users,
	 
	I am trying to use CHARMM27 taking the simulation conditions of two recent articles as guides in optimizing the simulation parameters. (DOI: 10.1021/ct900549r and DOI: 10.1021/jp101581h). 
	 
	-- Both are using PME for electrostatics and I am planning to do that too
	-- Both have rcoulomb = 1.2 (this value is optimal for CHARMM27 force field as mentioned)
	-- In the second paper, they have explicitly assigned rvdw = 1.2 (this value is also optimized for CHARMM27 force field) and they assigned the NS rlist = 1.2 because as I understand, using PME as a coulombtype requires rcoulomb to be equal to rlist. I have no problem following up until now and it seems that (rlist = rcoulomb = rvdw = 1.2) presents the best combination of cut-offs for using (coulombtype = PME and vdwtype = cut-off) along with CHARMM27
	 
	-- What I can't get is that in the first paper, they mentioned the following "van der Waals interactions were switched off between 1.0 to 1.2 nm"
	what does that mean in terms of cut-offs and vdwtype?
	Does that mean vdwtype = switch, vdw_switch = 1, rvdw = 1.2 ? 

yes. 

	
	and if this is what is meant, then rlist has to be 0.1-0.3 larger than rvdw to accomodate for the size of the charge groups as mentioned in the manual and accordingly, we can't keep rcoulomb = 1.2 because rcoulomb must be equal to rlist to allow using PME.

CHARMM FF doesn't use charge groups. Thus (as far as I understand) this warning can be ignored.  You still miss a few interactions because the neighbor-list only contains everything up to 1.2 thus an atom at a distance just above 1.2 and moves a bit inwards, it is not included until the next neighbor-list update. But because the interaction is already switched it is not very critical in my opinion. If you want to avoid this you can use PME-Switch but this is significant slower.

Roland 

	
	 
	 
	 

	--
	gmx-users mailing list    gmx-users at gromacs.org
	http://lists.gromacs.org/mailman/listinfo/gmx-users
	Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
	Please don't post (un)subscribe requests to the list. Use the
	www interface or send it to gmx-users-request at gromacs.org.
	Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
	




-- 
ORNL/UT Center for Molecular Biophysics cmb.ornl.gov <http://cmb.ornl.gov/> 
865-241-1537, ORNL PO BOX 2008 MS6309

-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 8308 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20101201/33a9100a/attachment.bin>