In Seurat, what data is VlnPlot() plotting?

Question

[Kristin Muench]

Hello,

I'm using the Seurat function VlnPlot() to visualize some of my data.

VlnPlot(object = data.combined, features.plot = c( 'Xist' )

When I plot it, the values range between 0 and 5.

I then wanted to extract the expression value matrix used to generate VlnPlot. By peeking inside the VlnPlot command, it looks like this is the command used to extract data:

exprs <- data.frame(FetchData(object = data.combined, vars.all = c('Xist' ) ), check.names = F)

However, when I run that, the values range between 0 and 120 - that is, they seem unscaled.

I see those same values no matter what I'm pulling for FetchData:

exprs <- data.frame(FetchData(object = data.combined, vars.all = c('Xist' ) ), use.imputed = T)
exprs <- data.frame(FetchData(object = data.combined, vars.all = c('Xist' ) ), use.scaled = T)
exprs <- data.frame(FetchData(object = data.combined, vars.all = c('Xist' ) ), use.raw = T)

What is VlnPlot doing differently to get data that only ranges between 0 through 5?

Thank you for your help.

Answer 1

[Kristin Muench]

UPDATE: It looks like the issue lied in confusing a combined Seurat object representing multiple samples and using an object that had a single sample. Previously, I had been plotting VlnPlot on combined sample, but accidentally trying to extract data from a single sample, which I'm guessing had not been scaled/regressed correctly, thus causing the confusion. Once I corrected this bug, this works as I would expect now:

exprs <- data.frame(FetchData(object = data.combined, vars.all = c('Xist' ))