search

aertslab / SCope

Fast visualization tool for large-scale and high dimensional single-cell data
GNU General Public License v3.0
68 stars 15 forks source link

User upload and loom download #515

Closed adamcc closed 1 year ago

adamcc commented 1 year ago

Thanks for making SCope available to the public, the core functions work well and are intuitive. I was able to identify an interesting cluster using the RBG text feature search. I have watched the tutorial videos.

The session page says "Datasets can be uploaded in user sessions using the left sidebar. Files should be in the loom file format and will appear under the "User Uploaded" category."

On the session that I am logged into, I can't see anything in the sidebar that looks like it could be a user upload function. When I drag a .loom file into the browser window, it just relaunches the same session.

It's not clear to me how to do follow-on analysis of a cluster, either in SCope in the browser or even which library I could use to filter some lassoed cells by gene expression.

I installed loompy to read the loom file format, which was able to load a test Cortex.loom sample file. Using a .loom downloaded from SCope yielded: does not appear to be a valid Loom file according to Loom spec version '3.0.0' Is there another package that would enable loom handling?

adamcc commented 1 year ago

A friend helped me sort this out: "The SCope website generates loom files using a unique implementation that deviates from standard loom. To access the SCope loom data with the python library, you'll need to use the SCopeLoomR package in R. Once the data is extracted, you can save it as an anndata object for use in scanpy."

https://github.com/aertslab/SCopeLoomR https://rdrr.io/github/aertslab/SCopeLoomR/