monocle3 clusters nCount thresholds automatically?

[R de Bruijn]

Hi everyone,

When I cluster my data part of my data is put in a low nCount cluster (below 3000) and a high nCount cluster (above 15500). Even though Seurat does not do this. Why is this? Seems almost like an intrinsic thing in the clustering algorithm of monocle3? I want to do a trajectory analysis. I feel like I should remove these cells if I do the trajectory analysis in monocle3 since the clustering seems based on the nCount instead of a biological gene expression reason. But the biologists feels like these are important cells and we should keep them in and she rather wants me to use the Seurat clusters since that package does not make these groups of cells into separate clusters but rather adds them to the other clusters. Also it's FACS sorted data so she things there won't be a lot of doublets. Also with doubletfinder I did not get this group out as doublets. Is there any way I can make monocle3 not put these cells in separate groups based on their nCounts?

Best,

R