cubi-tk snappy pull-sheets has undocumented samplehseet requirements

[Nicolai-vKuegelgen]

Is your feature request related to a problem? Please describe.

When trying to use snappy & cubi-tk snappy pull-sheets functions for projects not related to medical genetics other samplesheet templates can be more useful. However, it is not clearly documented which characteristics of the templates are actually necessary for cubi-tk snappy pull-sheets (and by extension snappy) to run without errors.

By trial and error I have now discovered the following constraints

• The 'Library Name' column in Sodar needs to contain a substring like '-N1-DNA1-WGS1' that identifies the library type. This reqiurement may be exclusive to the cubi-tk call and not actually affect snappy.
• The following source 'Characteristics' columns need to be present and - at least in partially - filled with specific non-empty values:
  • Disease Status (allowed: "affected", "carrier", "unaffected", "unknown")
  • Sex (allowed: "female", "male", "unknown")
  • Mother
  • Father
  • Batch

Describe the solution you'd like

Ideally a more flexible template that only requires the minimum information required for snappy to run should be added.

Describe alternatives you've considered

At a minimum these requirements should be documented such that they can be easily looked up.

Additional context
Snappy currently can't handle dashes in Source names: snappy-391
If this can not be fixed easily on the snappy side, cubi-tk snappy pull-sheets could give a warning in these cases.

[sellth]

Only Sex and Disease Status are required columns as they are not filled with default values in case of a no-show:

cubi-tk/cubi_tk/snappy/pull_sheets.py

Lines 304 to 318 in 3939f83

	row = [
	source.family or "FAM",
	source.source_name or ".",
	source.father or "0",
	source.mother or "0",
	MAPPING_SEX[source.sex.lower()],
	MAPPING_STATUS[source.affected.lower()],
	sample.library_type or "." if sample else ".",
	sample.folder_name or "." if sample else ".",
	"0" if source.batch_no is None else source.batch_no,
	".",
	str(project_uuid),
	sample.seq_platform or "." if sample else ".",
	sample.library_kit or "." if sample else ".",
	]

Easy fix would be to also set or "." / "0" for these values.

Also: Duplicate #120

[mbenary]

The column source.affected is only relevant for germline and thus does not exist in the somatic case. And cubi-tk snappy pull-sheets fails in line 310.