Enable generic & cancer dataset

[ericblanc20]

• Extends cubi_tk/snappy/pull_sheets.py to support datasets without Family, Father, Mother, Sex and/or Disease status as comments in the sample sheet.
• mRNA_seq also added to supported library types.
• The library name is used as default for the folder name when missing.

Note that the call to api.investigations.get is replaced by api.samplesheet.retrieve. I hope that it is correct...

[eudesbarbosa]

@ericblanc20, how did you test your changes? Is it something that could be used to stablish unit tests for SampleSheetBuilder?

[ericblanc20]

@eudesbarbosa I just checked it worked with a cancer dataset that I needed to download.

Re SampleSheetBuilder: I don't know. I am not knowledgeable enough with the cubi-tk, Biomedsheets and altamisa code to judge.

It seems to me that samplesheet logic is hard-encoded in the different instance of IsaTabVisitor (in add_ped, pull_raw_data & pull_sheets). Without an overview of the whole projects (cubi-tk, Biomedsheets, altamisa, snappy), I find it difficult to write something more general & I just add what I need with the least amount of disruption to the rest of the code. (I should really learn more about those...)

[eudesbarbosa]

@ericblanc20, implementing unit tests for this class doesn't look like a trivial task. I included a new ticket for it, #50.

[eudesbarbosa]

How did this affect the other sample sheet type?

@holtgrewe, care to comment? I'm not familiar with SODAR API syntax...

[eudesbarbosa]

@ericblanc20 and @holtgrewe , if the sodar-cli.api part is okay I suggest we just merge - this might even solve #52 and #53. I can take Friday to work on the unit tests, #50.