Identify perturbed atoms by index/number rather than by name

[lohedges]

I'm not sure if I'm missing something, but is there a reason why "name" is used to identify perturbed atoms in the SOMD pert file. Since this isn't unique it means that atoms must be renamed if there are duplicates, which can cause issues depending on how many atoms are perturbed. (The updated names must still adhere to the AMBER format, so there is a limit on their width.)

Instead, would it not be easier to identify atoms by their index or number, which are guaranteed to be unique? I understand that if this change was implemented then we would still need to support the old format (for backwards compatibility with FESetup). Perhaps we could increment the pert file version number and support both formats.

[jmichel80]

Hi Lester,

For a variety of reasons. If you use atom indices then the pert file only make sense together with a given topology, whereas if you match by names you can use the same pert file for different simulations (e.g. vacuum or solvated). It makes the pert file a bit more editable by hand or with simple scripts. Potentially this is less important if we move to a framework where the pert file is low level and hidden from the user. Will have to think about whether we should put efforts into the code change you propose.

---

Dr. Julien Michel, Senior Lecturer Room 263, School of Chemistry University of Edinburgh David Brewster road Edinburgh, EH9 3FJ United Kingdom phone: +44 (0)131 650 4797 http://www.julienmichel.net/

On Tue, Aug 13, 2019 at 10:31 AM Lester Hedges notifications@github.com<mailto:notifications@github.com> wrote:

Assigned #288https://github.com/michellab/Sire/issues/288 to @jmichel80https://github.com/jmichel80.

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHubhttps://github.com/michellab/Sire/issues/288?email_source=notifications&email_token=ACZN3ZG4V6CSTXQOR3MNESLQEJ5QPA5CNFSM4ILI4M22YY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOTAT7EDY#event-2552754703, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACZN3ZFOT2YQAWG5IBNDHFDQEJ5QPANCNFSM4ILI4M2Q.

The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

[lohedges]

But another SOMD restriction is that the perturbable molecule is always the first molecule in the system, right? (It needs to be renamed as LIG and placed first.) This means that it will be the first molecule that is loaded, so we would be able to identify the atoms by number/index just fine. (Perhaps there is another reason for the naming too.)

[jmichel80]

Hi,

This can be defined with the config keyword perturbed_resnum which is set to 1 by default. But the residue name has to match that used in the perturbation template. Technically we scan for the first molecule that contains residue number perturbed_resnum. The distinction between residues and molecule is another clunky aspect of the pert file format.

We could and should do better than that. I would like a more flexible framework to make it easier to setup perturbations in solvent or protein molecules but other urgent tasks have stopped this from happening.

---

Dr. Julien Michel, Senior Lecturer Room 263, School of Chemistry University of Edinburgh David Brewster road Edinburgh, EH9 3FJ United Kingdom phone: +44 (0)131 650 4797 http://www.julienmichel.net/

On Tue, Aug 13, 2019 at 11:51 AM Lester Hedges notifications@github.com wrote:

But another SOMD restriction is that the perturbable molecule is always the first molecule in the system, right? (You rename it LIG and place it first.) This mans that it will be the first molecule that is loaded, so we would be able to identify the atoms by number/index just fine. (Perhaps there is another reason for the naming too.)

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread. The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.