AMBER Archive (2004)

Subject: RE: AMBER: cut-off & MD

From: Ross Walker (
Date: Mon Dec 06 2004 - 15:17:30 CST

Dear Bo,

> I didn't set up the specific cut-off for the MD runs. So all the runs
> have a default cutoff setting at 8A. With my wide-type protein model,
> there are only significant conformation changes on that particular
> 'salt-bridge', but not breaking the H-bond. I am not sure if
> the cut-off
> values have some kinds of effect on it. I plan to try some
> replica runs
> with different cut-off values, to see if the links between the two
> residues break or not. Could you give me some comments if this is the
> reasonable way? Do the simulation time will be increased a lot with
> cut-off = 20 compared with cut-off =8?

If you ran with periodic boundaries using PME then the size of the cut off
is much less important than say a gas phase simulation since it only
truncates the van der Waals interactions in a PME simulation. These fall off
very quickly and so do not require a very large cut off to be treated
accurately. Thus in a periodic boundary pme simulation setting the cut off
larger is just going to take a lot longer with only a very very small
increase in accuracy. Note a 20A cutoff will take significanltly longer than
an 8A cut off since the number of pairwise interactions will increase

Note: if you are running a gas phase simulation then the system will be very
sensitive to cutoff. You will still have significant error even out to say
16A cutoff.

I hope this helps.
All the best

|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- |
| | PGP Key available on request |

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to