I am planning to use velocyto to quantify RNA velocity on my samples and I have been doing some reading but I am a bit confused about what the input to velocyto should be. From what I understood I can use the BAM files generated from Cellranger in velocyto and then parse the aligned reads into a spliced and unspliced matrix. However, when generating the BAM files do I need to include --include-introns in the cellranger count command?
Velocyto has generally not worked well for me (high memory useage and long run times). As an alternative I've taken to STARsolo which can additionally output Velocyto-like counts, and as an alternative the alevin-fry RNA-velocity tutorial. Alevin-fry is the scRNA-seq version of Salmon.
The default settings should be fine as CellRanger will note which reads map to exons and which to introns in the RE tag of the reads in the resulting BAM file (see here)
The naming of the parameter is a bit unfortunate, you can read more about the alignment step here:
Cell Ranger uses an aligner called STAR, which performs splicing-aware alignment of reads to the genome. Cell Ranger then uses the transcript annotation GTF to bucket the reads into exonic, intronic, and intergenic, and by whether the reads align (confidently) to the genome. A read is exonic if at least 50% of it intersects an exon, intronic if it is non-exonic and intersects an intron, and intergenic otherwise.
I think that the --include-introns option mostly makes a difference for the matrix generation of UMI counts, but it's definitely not spelled out well enough to grasp that unless you've read the documentation many times.
To count these intronic reads, the cellranger count and cellranger multi pipelines can be run with the option include-intron
Velocyto has generally not worked well for me (high memory useage and long run times). As an alternative I've taken to STARsolo which can additionally output Velocyto-like counts, and as an alternative the alevin-fry RNA-velocity tutorial. Alevin-fry is the scRNA-seq version of Salmon.