-
Notifications
You must be signed in to change notification settings - Fork 0
7_Submit single cell data to GEO
Olivia Waltner edited this page Dec 26, 2023
·
4 revisions
-
Wherever your experiment is located, create a 'geo_upload' folder
-
Prepare single cell data
in R:
meta<-seu@meta.data
names(meta) # double check that you want to keep all the columns of your metadata, if not subset columns above
#add umap coordinates to metadata
meta$UMAP_1<-seu@reductions$umap@cell.embeddings[,1]
meta$UMAP_2<-seu@reductions$umap@cell.embeddings[,2]
write.csv(meta, file.path( "geo_upload/metadata.csv"))
- Export processed count matrices, if you have more assays (ADT,ATAC, etc) create separate files
sct_mat<-seu@assays$SCT@data %>% as.sparse()
write.csv(sct_mat, file.path("geo_upload/rna_counts_processed.csv"))
Copy cellranger bams or fastqs of sample runs to folder, make sure they have unique file names.
If you have something like this, then your folder is ready!
- Go here: https://www.ncbi.nlm.nih.gov/geo/info/seq.html and click on
It will direct you to making an NCBI account if you don't already have one
5)Fill out your profile. Here is an example. Hit 'save'.

- Then go back a page and hit 'transfer files' again. Something like this should pop up.
find this

and this

- Edit these lines of code, then run on terminal/bash on r markdown
ncftp
set passive on
set so-bufsize 33554432
#make sure below line is geoftp:PASSWORD FROM ABOVE@ftp-private.ncbi.nlm.nih.gov
#so for me it looks like
open ftp://geoftp:inAlwokhodAbnib5@ftp-private.ncbi.nlm.nih.gov
#change this to your directory found in step 6
cd uploads/owaltner@orcid_MlmJa2eM
#change this to the path to your geo_upload directory
put -R /fh/fast/furlan_s/.../geo_upload
- Nagivate to geo_upload directory and get Md5 checksums
cd /fh/fast/furlan_s/.../geo_upload
md5sum rna_counts_processed.csv
#i got this string: f2a01b9c6ee751a97ddeda99301da43a
md5sum metadata.csv
#returned this: edfeaa082abe9797b4033be75dd905da
- Insert md5sum values in sheet 2 of GEO metadata excel spreadsheet
- navigate to GEO submission page and select: