add option to keep original downloaded silva files and/or different taxonomy files?

[mikerobeson]

Should we allow an option to keep the raw imported SILVA files we download?

My thinking is that if users would like to construct classifiers with and without species labels they will not have to go through the download option twice.

Or... given that this is a nice convenience method, we can simply return both taxonomy files, i.e., with and without species labels, and then there'd be no need for the option of writing out the raw files? This would will make constructing the silva dbs much easier for each qiime release.

I suppose we can do it all.. but that might be confusing.

[nbokulich]

I agree downloading twice is an annoyance...

do we have any methods that could just slice the taxonomy? Maybe we should add one. A method to rename taxonomic labels (by slicing, removing rank handles, etc) would actually be quite useful, e.g., for manually modifying taxonomy labels before or after classification.

[mikerobeson]

Not that I am aware. I know we previously discussed being able to parse / modify the taxonomy files. I suppose we can simply do that. Would definitely be much faster to slice and dice the taxonomy rather than re-processing.

[nbokulich]

okay let's go with slicing and dicing. it would be a fun function to write too! do you want to take this one or should I?

[mikerobeson]

Well, I do need more pandas practice so I can try tackling this one. Unless you think there is urgency to do this ASAP. :-)

[nbokulich]

no urgency on this IMO... this will be a fun one for you 😄

[thermokarst]

I love this idea. We have been talking about a cache-type system in QIIME 2 for a while. I think there are some issues floating around on qiime2/qiime2...