Closed XinWang19 closed 2 years ago
Dear @XinWang19 ,
This issue was addressed in https://www.nature.com/articles/s41596-020-0336-2 . At the moment the solution we recommend is to run SCENIC several times to assess the stability of the results. This is automated in VSN-pipelines (i.e. see "Scenic multiruns").
I hope this helps
Hi,
I ran SCENIC on my single cell dataset (about 900 cells) twice, with a difference of 14 cells. However, I found there were considerable amount of regulons differing after I added these 14 cells (as following). How could I ensure the reproduciblity between the datasets? Thanks!
Xin