Integrated Seurat object change name of the two conditions

Hi,

I have a scRNAseq integrated seurat object composed by my control and my treatment. On this object I did all the analysis and I have all the plots needed (umap etc). Now it has been asked to use different names and instead of "control" and "treatment", I should use two different names (let's say "A" and "B" where A is my "control" and B is my treatment) for the plots showing the comparison of the two conditions (eg. violin plot).

So is there a way I can change "control" and "treatment" with "A" and "B" on my seurat object (maybe creating another column that I can refer to when I use for split.by) without have to re-do the integration from the beginning? or shall I just give up and modify the labelling in photoshop?

Thank you

Camilla

r integrated seurat

• 3,579 views • link

written 2.9 years ago by camillab. •

Hi, See this prev. post where I contributed to, on accessing the meta data slot in a seurat object. There is more than one way.

• 0 views • link

written 2.9 years ago by Amitm

thanks but my metadata looks different from yours. Will it work even there are not the same "structure"?

                           orig.ident nCount_RNA nFeature_RNA percent.mt  sample
Control_AAACCCAAGTACAGCG-1    Control       4133         1921   3.629325 Control
Control_AAACCCAAGTTAGTAG-1    Control       1234          700   6.320908 Control
Control_AAACCCACAAAGCAAT-1    Control       3361         1630  11.454924 Control
                           integrated_snn_res.0.2 seurat_clusters integrated_snn_res.0.3
Control_AAACCCAAGTACAGCG-1                      8              12                     12
Control_AAACCCAAGTTAGTAG-1                     12              11                     11
Control_AAACCCACAAAGCAAT-1                     14              15                     15

• 0 views • link

written 2.9 years ago by camillab. •

Yes, its similar structure. This snippet should add a new metadata column "sample.New" with what you want to do

Assuming a seurat obj. called "my.seuObj".

Check the distrib. of your labels in the pre-existing metadata column

table(my.seuObj@meta.data$sample)

Create a metadata column with dummy value

my.seuObj@meta.data$sample.New <- "placeholder"

Check

table(my.seuObj@meta.data$sample.New)

Change the value in "sample.New" metadata column w.r.t "sample"

my.seuObj@meta.data$sample.New[which(my.seuObj@meta.data$sample == "Control")] <- "A" my.seuObj@meta.data$sample.New[which(my.seuObj@meta.data$sample == "treatment")] <- "B"

Check

table(my.seuObj@meta.data$sample.New)

Check 2. This should be all "Control" value

my.seuObj@meta.data$sample[which(my.seuObj@meta.data$sample.New == "A")]

• 0 views • link

written 2.9 years ago by Amitm

great! thank you very much!

• 0 views • link

written 2.9 years ago by camillab. •

0 answers

No answers yet.