Question

Flow Cytometry Data Analysis by Seurat

0

Entering edit mode

4 weeks ago

gdfsnkfns • 0

Hello everyone.

I'm now trying to analyze relatively big FACS data (about 500 samples, 16 colors). I think Seurat is suitable for our purpose (e.g. conducting dimensional reduction of the data such as tSNE), but does it make sense?

I have already analyzed FACS data using Flowjo (manually gated and grouped cells with cell surface markers as indicators), and prepared matrix table summarizing the percentage of cells in each cluster for each sample. (e.g. PD-1+Tim3- in CD8+: x% PD-1+Tim3+ in CD8+: xx% PD-1-Tim3+ in CD8+: xxx% PD-1-Tim3- in CD8+: xxxx% ....... for each sample)

I suppose it would not be a wrong way to use the above data matrix as input and conduct the standard Seurat workflow without normalization process. What do you think?

I would be happy to hear your opinion.

Best regards.

Seurat • 227 views

ADD COMMENT • link updated 4 weeks ago by Ram 43k • written 4 weeks ago by gdfsnkfns • 0

1

Entering edit mode

I think Seurat is suitable for our purpose (e.g. conducting dimensional reduction of the data such as tSNE), but does it make sense?

No, Seurat expects counts, not continuous intensities. Normalization and data transformation is different in FACS than in NGS. There is dedicated software for FACS analysis. FlowJo is probably the most common one, but there is also lots of packages for flow analysis at Bioconductor, for example FlowCore. These packages have guided tutorials to follow. Don't reinvent the wheel.