search

marrink-lab / vermouth-martinize

Describe and apply transformation on molecular structures and topologies
Apache License 2.0
86 stars 38 forks source link

merge eunit and region flag #465

Closed fgrunewald closed 1 year ago

fgrunewald commented 1 year ago

This fixes #459 at least partially. As stated in the issue there is still some ambiguity as to what exactly should happen if a resid region is defined for a molecule with more than one chain and same resids per chain

fgrunewald commented 1 year ago

As far as the behaviour goes the eunit=chain is really just an edgy edge case. It only comes to be important when two chains, which are labelled as such are connected by a disulphide bridge. Any thing else will end as two separate molecules or one molecule. Thus I think it is fine having the eunit flag being either all, molecule, chain, or residrange.

Added bonus: The behaviour is much more understandable, because almost all people dealing in proteins consider molecule=chain.

pckroon commented 1 year ago

Then maybe chain + region was a poorly chosen example. How about molecule + region? You may want an elastic network only within a molecule/chain, but only between resids 200-400. So if you have 2 homodimers that are close together you only want EN between resid 200-400, but not between the two dimers. I think that's not possible with the current CLI suggestion

fgrunewald commented 1 year ago

@Tsjerk any thoughts on this

fgrunewald commented 1 year ago

@pckroon I suggest we roll with this PR for now, so that we can put out a preprint. At least the current PR is better than the buggy behaviour form before.