search

ktrns / scrnaseq

Workflow for single-cell RNA-seq analysis using Seurat
MIT License
37 stars 15 forks source link

DEGs contrast in separate script #100

Open kosankem opened 3 years ago

kosankem commented 3 years ago

Persönlich fand ich es sinnvoller und effektiver die DEGs contrast Analyse separat nach erstmaligem Durchlauf des Skripts einzustellen und zu starten. Praktisch heißt das, ich musste das Skript zunächst einmal laufen lassen, dann die Parameter für die DEGs eingeben und das Skript erneut laufen lassen, um die gewünschten Darstellungen erhalten zu können. Nach wie vor möchten wir gerne am Konzept festhalten, so nah wie möglich an einen guten ersten Standard-Report (bei einmaligem Anstoßen des Skriptes) zu kommen, welcher die wichtigsten Ergebnisse zur Studie bereits bestmöglich enthält und dann die Basis für eventuelle vertiefte und projektspezifische Analysen darstellt.

ktrns commented 3 years ago

Yeah, I'd be open for that. Let's discuss it with @andpet0101.

Sometimes the clients know upfront that they want to compare sample1 and sample2, but whenever clusters are involved, the workflow currently needs a complete re-run. On the upside, currently the whole analysis is included in a single report.

ktrns commented 2 years ago

Hi @kosankem,

In order to finish our first major release, I suggest that we keep this discussion open, and work it out after the release, as it is not super urgent. Is that alright for you?

Best wishes Katrin

kosankem commented 2 years ago

Yes, that is fine for me. Thanks.

kosankem commented 2 years ago

Another small command to the DEGs:
We were having a problem for a while to run the deg_contrast analysis. We got the error:

Error in match(x, table, nomatch = 0L) :
'match' requires vector arguments

I could finally solve it by specifying in the functions_degs.R script line 461 the package from which the Assays() function is called:

if (contrast[["assay"]] %in% Seurat::Assays(sc)) {

It seems that by default Assays() of the SummarizedExperiment package was used in my case. Although this is mainly an issue of the setup of our R packages, the specification “Seurat::” might be helpful.