PCA analysis

It is useful to include a PCA analysis for detection of outliers or general agreement of samples. The idea is that this is incorporated in the MultiQC report.

added Future + 1 deleted label

In principle, we already have a script here: https://git.scicore.unibas.ch/zavolan_group/pipelines/zarp/-/blob/dev/scripts/perform_PCA.py

So what needs to be done is:

• Create a repository with the script under https://github.com/zavolanlab (like https://github.com/zavolanlab/merge_kallisto or https://github.com/zavolanlab/zgtf)
• Add CI (e.g. github actions)
• Add unit tests (ideally pytest)
• Create slim Dockerfile
• Integrate the script as a snakemake rule(s) (e.g. kallisto genes/transcripts, salmon genes/transcripts)
• Remove perform_PCA.py script from scripts directory

Of course, we can also use another tool (if already out there)

assigned to @gypas

changed milestone to %pipeline

changed milestone to %downstream

@zavolan Do you want to run the PCA analysis only for genes or also for transcripts?

mentioned in merge request !77 (merged)

mentioned in issue #142 (closed)

mentioned in commit c535a892

Fixed in !77 (merged)

closed

removed Future label