search

ratschlab / scim

Code for Universal Single-Cell Matching with Unpaired Feature Sets
GNU Lesser General Public License v3.0
20 stars 3 forks source link

Questions about cell subtypes #8

Open 5rycaa opened 3 years ago

5rycaa commented 3 years ago

Good morning!

In your integration of melanoma patient sample of B and T cells, I was wondering was there a specific reason behind going for B/T integration? More specifically, have you tried in any of your examples to integrate a bit more complex cell subtypes eg. CD4 and CD8 cells (known to be transcriptomically challenging to resolve)?

And if you don’t mind me asking, have you attempted to integrate multiple samples at the same time?

Thanks a lot!

stefangstark commented 3 years ago

Hi,

We are currently working on integrating more complex subtypes and e.g. tumor cells. Multiple samples are also in the pipeline. Depending on what assumptions you would like to make about your cohort & technologies the alignment problem can become quite complicated. Since you need to align the samples in addition to the cell types & subtypes.