Hi,
first of all, really beautiful work that you guys do! Thanks for all the effort and for sharing with the community.
I am trying to reproduce the pancreas analysis from scratch (downloaded the fastq and run cell ranger & velocyto).
With all 4 files (E12.5-E14.5) I end up already with more than 40000 cells.
In your dataset (scv.datasets.pancreas()) there are only about 3600 cells.
Could you share the way of filtering you did (with cellranger or scanpy)?
Moreover the phase plot of Cpe looks about right, but the velocity representation doesn’t really fit:
I would be glad for any advice.
Best, Sandra