[Feature]: Add Kallisto + BUSTools Workflow Parts

[apeltzer]

This PR is intended to serve as a WIP to add Kallisto + Bustools to the scrnaseq pipeline.

PR checklist

• This comment contains a description of changes (with reason)
• If you've fixed a bug or added code that should be tested, add tests!
• If necessary, also make a PR on the nf-core/scrnaseq branch on the nf-core/test-datasets repo
• Ensure the test suite passes (nextflow run . -profile test,docker).
• Make sure your code lints (nf-core lint .).
• Documentation in docs is updated
• CHANGELOG.md is updated
• README.md is updated

Learn more about contributing: https://github.com/nf-core/scrnaseq/tree/master/.github/CONTRIBUTING.md

[apeltzer]

Downstream analysis can be done using: https://scanpy.readthedocs.io/en/latest/index.html

I'm following this tutorial here :

https://www.kallistobus.tools/getting_started.html

Details about some steps (e.g. gene_map creation as implemented now in this PR) are explained here:

https://www.kallistobus.tools/getting_started_explained.html

[olgabot]

Oh oops I had worked on this too but forgot to submit a PR :(

https://github.com/czbiohub/scrnaseq/tree/olgabot/kallisto-implementation

Should I make a PR to this one with my changes?

The V3 chemistry is compatible with V2, so we can use that as default. I also decided to match kallisto's technology naming to make it easier to grab the whitelists and support multiple technologies.

[apeltzer]

Páll advised against using the V3 whitelist in our case as the correction needs to much memory to complete. I filed an error report in the BUSTools repository for that: BUStools/bustools#6 , where he replied with the comment to rely on v2 whitelists at least for the tests. Didn't help, unfortunately yet, even when skipping the correction.

We can merge this into kallisto-bustools here and you can then also open PRs against that one - however, the tests aren't working yet which is maybe the first thing we should address...

[olgabot]

Ohhh interesting I didn't realize that about the V3 whitelists. I remember them being a lot larger so it makes sense that it's not recommended for testing.