Hi,
There's a lot to unwrap here (forgive the pun) but I'll do my best to
answer all of your questions.
On Tue, Nov 10, 2020 at 6:13 AM Gustaf Olsson <gustaf.olsson.lnu.se> wrote:
>
> 1) When would it be important to use the “noimage” option for radial analyses? I see the options though not the motivation and I have not been able to generate a difference larger than 0.001 with my trial sets
Realistically, never. Not imaging the distances that go into the RDF
in a system with PBC is not formally correct. The one difference you
will notice here is calculation speed, since using the minimum image
convention has a not insignificant cost associated with it. If you're
doing something like a water oxygen RDF where you've got tons and tons
of averaging, the differences will be slight. In cases where there are
fewer entities to average over the differences will become more
pronounced. So the keyword is there as a short-cut, but there's a
reason imaging is always on when PBC are present.
One question - the .001 difference you mentioned for RDF was between
image and noimage right? Not image and autoimage?
> 2) Is there any preference for using autoimage before performing hbond analysis or using the image option during the analysis.
>
So unlike the RDF, for the 'hbond' command' imaging of distances is
*not* on by default. This is why you will see big differences between
image/no imaging. Imaging is off because in many cases for 'hbond' to
work right you don't really need it. If you're just looking at
solute/solute hydrogen bonds for a single molecule, imaging is not
needed. If you've done your imaging properly (e.g. via autoimage) in a
well-solvated system so that there is a larger buffer of bulk water
around your solute and no imaging artifacts exist, again imaging is
not needed. However, if your system is more condensed you may indeed
need imaging. So really it depends on your system.
> Considering both efficiency and accuracy. Performing longer, large system simulations often produces large amounts of data. From an efficiency point of view, if there is a need to “autoimage” trajectories before analysing, the time consumption for this action and the extra disk space required may indicate using iwrap=1 during simulation being more efficient for these reasons.
The problem here is that iwrap=1 does not image in a "smart" way - I
think imaging is done based on the location of the first atom, so you
can end up with imaging artifacts (like splitting DNA strands) - this
was the whole reason 'autoimage' was even developed.
> I have no idea regarding accuracy/performance hit of using the "noimage” option with radial so I donated know if this would outweigh the benefits.
>
> How would this compare to using iwrap=0 and image option in hbond with regard to speed/accuracy?
Even if you image a trajectory before running hbond (or RDF for that
matter), it's not the same as having imaging on in the command itself.
As I said before, for 'hbond' it may _appear_ the same if none of your
hydrogen bonds are across the periodic boundary, but otherwise image
and noimage are two very different analysis modes.
>
> Is there potentially any nice overview indicating which analyses could be affected by wrapping and/or recommendations for settings not listed with the particular analyses in the manual?
Not really, although this would probably be nice to add to the Actions
table. In general if you search for 'image' in the cpptraj manual that
will give you an idea:
https://jenkins.jasonswails.com/job/amber-github/job/cpptraj/job/master/lastSuccessfulBuild/artifact/doc/CpptrajManual.pdf
Hope this helps a bit,
-Dan
>
> Best regards
> // Gustaf
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Nov 10 2020 - 06:30:02 PST
