search

AlexsLemonade / alsf-scpca

Management and analysis tools for ALSF Single-cell Pediatric Cancer Atlas data.
BSD 3-Clause "New" or "Revised" License
0 stars 1 forks source link

Single Cell Data QC #43

Closed jashapiro closed 3 years ago

jashapiro commented 4 years ago

The current Alevin workflow does not perform any particular QC steps; we have been doing this by hand so far.

We should implement some automated QC!

The simplest version of this would be adding alevinQC to the run-alevin.nf workflow. For this to work, the current alevin workflow will need to be modified to include the --dumpFeatures flag

While this is pretty easy, it isn't always as robust as we might like, and does not support cross-library summarization (like MultiQC?), which we might like to add in the future.

To allow for future flexibility, we should make sure to output the entire AlevinQC report directory, not just the html files, so that we may be able to process them later after reading in data tables with readAlevinQC() or similar.

jaclyn-taroni commented 3 years ago

miQC: https://www.biorxiv.org/content/10.1101/2021.03.03.433798v1

cgreene commented 3 years ago

Could be useful for training too. It takes the challenge of picking a threshold for dead cell / debris removal and reframes it as a mixture model. The cases where it doesn't work well are samples where there is not a mixture (i.e., no dead cells/debris or no live ones).

jaclyn-taroni commented 3 years ago

Can this be closed?

jashapiro commented 3 years ago

Yes, I suppose it is largely superseded by our current QC reports.

allyhawkins commented 3 years ago

I would say yes.