search

alexQiSong / scSTEM

Single cell STEM (scSTEM) is a shiny app based R package for visualizing and clustering genes in pseudotime ordered single cell RNA-seq data. scSTEM is a GUI based tool and thus does not require any coding experience.
18 stars 2 forks source link

use of Seurat object #4

Open gebdu opened 1 year ago

gebdu commented 1 year ago

Hi,

can I use Seurat object (with UMAP/clustering already done) just to do the trajectory analysis ?

thanks

alexQiSong commented 1 year ago

Hi,

can I use Seurat object (with UMAP/clustering already done) just to do the trajectory analysis ?

thanks

Hi Gebdu,

Currently loading from Seurat object is not supported yet. But this could be a useful extension. I would consider adding this functionality soon. Thanks for providing the feedback!

Thanks, Qi

Dalhte commented 2 weeks ago

Hello there I extracted the info from my seurat object like this :

write expression counts matrix

library(Matrix) counts_matrix <- LayerData(object = obj[['RNA']]) writeMM(counts_matrix, file='/counts.mtx')

write gene names

write.table( data.frame('gene'=rownames(counts_matrix)),file='/gene_names.csv', quote=F,row.names=F,col.names=F )

For cell_meta.csv:

I runned velocyto, extracted the anndata obs and variable, and from the obs.csv, keep only the barcodes and the pseudotime index

Cannot say if it is working yet. The files were downloaded successfully, the UMAP reduction is nice, and the inference of trajectory still running (it has been 3 hours now :p)

Best

David

Dalhte commented 1 week ago

Just for you to know that what I proposed above is working in my hands Best David