How should we organize this?
First of all we should have --final in any result that is preprocessing. But what default should it have? Also preprocessing Results should expose what term they add to preproc. For Trimmomatic it would be tmtic_params.
TRUE
If the user requests just one preprocessing in command - everyrhing is fine, we just mark that this preprocessing is final for this samples, and the user can continue. But if the user wants to request multiple preprocessings? They all will be set as final in order of appearance in request, so the last one requested will end up being final.
Another issue is what to if the user requests preprocessing that is doable but not created? Say, have just raw samples at the moment and user requestsmetaphlan2 -d <DATASET> -p raw__tmtic_def. We should first check that it is doable, and then add new result to global requested results. Marking resulting files as final seems pretty logical.
Chaining preprocs and sample sets in general? assnake result trimmomatic -d DF rm-human
How should we organize this?
First of all we should have
--finalin any result that is preprocessing. But what default should it have? Also preprocessing Results should expose what term they add topreproc. For Trimmomatic it would betmtic_params.TRUEIf the user requests just one preprocessing in command - everyrhing is fine, we just mark that this preprocessing is final for this samples, and the user can continue. But if the user wants to request multiple preprocessings? They all will be set as final in order of appearance in request, so the last one requested will end up being final.
Another issue is what to if the user requests preprocessing that is doable but not created? Say, have just
rawsamples at the moment and user requestsmetaphlan2 -d <DATASET> -p raw__tmtic_def. We should first check that it is doable, and then add new result to global requested results. Marking resulting files as final seems pretty logical.Chaining preprocs and sample sets in general?
assnake result trimmomatic -d DF rm-human