Closed fgrunewald closed 1 year ago
@pckroon I'd prefer we get rid of one of the flags either rb_region
or rb_unit
and for the reminaing flag we accepts either the keywords chain
, molecule
, all
or a resid region
That sounds like a reasonable solution. It makes the argparse part a bit more complicated due to argument validation, but not too much I think
well the rb_region flag already takes both keywords and also resid lists at the moment
@pckroon This is actually a bit more complicated than I thought: What should happen when we have one molecule consisting of two chains and define a resid region? Should the EN be formed between the chains or not?
Logically all criteria should get ANDed. For now though, the CLI is limited to providing just a single criterion (right?). So yes, it should be formed between the chains.
Someone messed up when implementing #422. The lines below are actually non-sense because they overwrite the e-unit criterion.
https://github.com/marrink-lab/vermouth-martinize/blob/aa53b3e3ca734032518bd56b3b8dad0831653612/bin/martinize2#L737-L741