welch-lab / MultiVelo

Multi-omic velocity inference
BSD 3-Clause "New" or "Revised" License
105 stars 12 forks source link

Velocyto .loom file creation #38

Open AlessiaLeclercq opened 9 months ago

AlessiaLeclercq commented 9 months ago

Hello, I'm trying to reproduce the .loom file for the 10X embryonic E18 Mouse Brain using velocyto CLI and compare it with the one you provide. I run the 10X command on the folder containing the "filtered_feature_bc_matrix" and the "GEX position-sorted alignments BAM" (from cellranger arc output) using the mm10 reference .gtf file. However, when I try to compare the resulting .loom file with yours (using scanpy) the annotated data objects have different results for the X matrix and the spliced/unspliced layers. Did you use any expressed repeats annotation in the CLI command?

danielee0707 commented 9 months ago

Hi, we didn't use expressed repeat masks when running Velocyto.