Simplify output_name

[Qrouger]

Input File :
P09783,3;P0A3W6,3;P0A3W4

Old name output :
P09783_and_P09783_and_P09783_and_P0A3W6_and_P0A3W6_and_P0A3W6_and_P0A3W4

New name output :
P09783_homo_3er_and_P0A3W6_homo_3er_and_P0A3W4

[jkosinski]

Thanks again! A shorter naming convention would be useful. Did you check whether the analysis pipeline still works with these names? Also, are you using the script mode or snakemake pipeline?

[Qrouger]

I use it only on script mode, the singularity image seems to work.
predictions_with_good_interpae.csv

[DimaMolod]

would this work if is the delimiter is other than '+'?

[Qrouger]

Between differents proteins, delimiter is always "+", if there is no "+" in self.input[0] so list_oligo will look like [P13979] (monomer) or [P13979:2] (dimere)

[DimaMolod]

Sorry, I mean we can choose a delimiter when run structure predictions. The '+' symbol is the default one, but it can be set to any other character:
https://github.com/KosinskiLab/AlphaPulldown/blob/main/alphapulldown/scripts/run_structure_prediction.py#L153
Therefore we have to pass the delimiter flag as well, otherwise it will not work for any custom delimiters other than '+'.
Then the code would look like:
list_oligo = self.input[0].split(self.delimiter)
and we need to pass this delimiter beforehand e.g.
object_to_model = MultimericObject( interactors=interactors, delimiter=delimiter, ...
On the other hand, maybe it makes more sense to move all this logic from objects.py to modelling_setup.py and keep MultimericObject logic as simple as possible (given it's already far from being simple)

[Qrouger]

Ok, sorry I don't know delimiter flag exist. Yes that make more sense to add this at modelling_setup.py.