Seurat scRNAseq analysis: what is the mito gene list

Question

[wenbinm]

Hi there, I am trying to analyze 10X genomics output with Seurat. After reading in data using "Read10X" function and create an Seurat object, I am stuck at this step:

> mito.genes <- grep(pattern = "^MT-", x = rownames(x = my_data@data), value = TRUE)

For the data I have, row names of the expression matrix are gene names ("Sox17" "Mrpl15" "Lypla1" "Tcea1" "Rgs20" etc) or ensembl ID, where no genes begin with "MT-".

Does anyone know the mitochondria gene list Seurat uses for quality control? Or I should generate a mitochondria gene list on my own? I am working on mouse samples.

Many thanks for your help!

[Arup Ghosh]

Mouse mt gene ids will start with "mt-"

[Lauren]

is rat mt gene ids also "mt-"?

[Lauren]

I was told it is ^Mt for rat

Answer 1

[igor]

Looks like you have a mouse genome. Instead of ^MT-, you can look for ^Mt- or just add ignore.case = TRUE to the grep.

[wenbinm]

It works. Thank you! It seems that Seurat is always using these 13 genes: 'Nd1','Nd2','Co1','Co2','Atp8','Atp6','Co3','Nd3','Nd4l','Nd4','Nd5','Nd6','Cytb'

Answer 2

[rogangrant]

At least in my case, mouse mitochondrial genes are flagged with ^mt-, rather than ^Mt-. But either will still work if you set ignore.case = TRUE as per igor's suggestion.