Enhancement: Leave structural ions in file

[KasperBuskPedersen]

Dear Martinize2 devs

Structural ions are quite common, however martinize2 currently deletes all ions from the input pdb so the user need to add them back in. A nice quality of life fix would be to detect ions in the input file and just leave them be in the output or have a flag that toggle this behavior if the current is intended.

Best regards Kasper

[KasperBuskPedersen]

Oh, maybe this can just be done with the -ignore flag according #74. Sorry if that is the case. Could possibly add a little more text to the -h documentation about this :) Cheers Kasper

[pckroon]

Hmmn I can't reproduce this I think. On the master branch we don't drop/ignore any residues by default. Running martinize2 -f 1ubq.pdb produces a lot (116) of warning about the HOH residues for which no datafiles are available. So to preserve the ions and solvent you'd need to add the data files (as if adding a new residue). The mapping to Martini is going to be hard/impossible though, since multiple molecules will correspond to one bead... I played around with it, but never finished anything along those lines.

I'm not sure how to improve the --help message, I'm open to suggestions.

[fgrunewald]

@KasperBuskPedersen do you want to keep the ions in the output? We could do this quite easily if the mapping files are provided. That's something you could even do yourself and the contribute to martinize2. ;)

[KasperBuskPedersen]

@fgrunewald Yes exactly, and maybe some control over which should stay (i dont know how this would be best done). For example, you demonstrated a high-througput set up in the martinize2 pre-print where you could (in principle) simulate thousands of proteins. However, so many would have structural ions, which means that a large part of those simulations would be possible wrong without having these crucial ions - it is only because that elastic networks hide these problems right now, but with increased accuracy in the protein model, these things will eventually show.

All the common structural ions will just be a 1-1 mapping to either a Q or D bead, so super simple. I'll note it down for when im done with the PhD and i might come back.

Cheers Kasper

[yummy-hat]

Hello,

Has anyone thought more about this issue? I have several structural ions in my protein, but I'm not sure how to adjust the positions of the CG ions.

I have some calcium ions that are being coordinated. Is there a strategy to try to maintain that coordination in Martini3?

[pckroon]

Not actively. I'm not sure what the state of the model is in this regard (I'm not part of the Marrink lab any more :'( ), but I am pretty sure we don't have the mapping files yet. PR welcome! :)