Hello everyone, I have a scRNA-seq dataset with control, condition 1 and condition 1 + 2 (sh). Condition 1 induces a phenotype that has been previously described and I was able to characterize it by adding some details given the higher number of cells in our dataset compared to previous works.
Condition 2 is expected to influence the "normal" phenotypes induced by condition 1.
I was suggested to perform an exploratory evaluation of the clusters I can recognize in condition 1 and in condition 2 alone. Unfortunately, I am not able to recognize dramatic changes. I know that the system that induces condition 2 works since differences can be detected in bulk seq.
How should I approach this analysis? I have seen approaches such as pseudo-bulk analysis, but I fear I have too few samples for having a robust estimation of dispersion (2 for each sample).
Thank you
Do you have 2 biological or technical replicates for each condition? Also, you mentioned you didn't find dramatic changes with your analysis. What did you do for that analysis?
Hi! Two biological replicates. The previous analysis has been performed on Seurat.
"performed on Seurat" is like saying that the repair was done with a hammer and skrewdriver. Seurat has plenty of functions, you need to be a bit more specific.
Hi, I analyzed the datasets separately as follows:
I applied the standard worklow looking for markers using WSRT (find all markers).
The first dataset has been extensively characterized, I also compared DEGs between subpopulations and performed ORA and GSEA.
In the second I specifically searched for a subpopulation that I expected to be deregulated by condition 2. However, I still can find its markers.
I was wondering if there are more appropriate solutions to check if there are quantitative differences.