Removed ReferenceGenome as an option for Capture_Type

[nevrome]

As proposed in #66.

A special Date_Type is imho not necessary. Reference genomes conceptually exist outside of space and time. And multiple other .janno columns also don't apply for them. All of these can equally be set to n/a.

[stschiff]

OK, but there should be some way to indicate that a "sample" is actually a reference genome. If not in Date_Type, then where do we indicate that?

[stschiff]

Fine, after Discussion on March 19.

[nevrome]

If I remember our discussion correctly, then the argument against a special Date_Type value for reference genomes was that we would also need special values for various other columns. Dating is just one aspect affected.

For the time being I think it's fine to set everything to n/a that does not apply for reference genomes. Maybe eventually a field for the overarching Sample_Type would be useful.

[nevrome]

Probably it's not sensible to merge this change to Capture_Type as long as we have no alternative.

What do you think about a Sample_Type column, Stephan? The difficulty is probably to come up with meaningful categories. Maybe just three simple classes?

• AncientDNA
• ModernDNA
• ReferenceGenome

Each of these subsume various source materials and data generation techniques. We could inform such a column from the Date_Type column, but we would not have to rely on it any longer for this major distinction.

[stschiff]

I happen to be convinced that it's OK to drop this. You correctly pointed out that it's easiest to just exclude those packages with reference genomes and that's that. So I think it's OK to just go ahead with this PR