BAM to BED step is the slowest, so not sure we need to always run it?
Also multiQC report currently is off by default, and would require manually editing the snakemake input requirement. Also, with UG R1-only mode, adding the report req triggers Snakemake trying to run R1, R2, pair before summarizing the final report. Alternatively we just remove it altogether?

From 100 million UG 10x reads on cluster mode
BAM to BED step is the slowest, so not sure we need to always run it?
Also multiQC report currently is off by default, and would require manually editing the snakemake input requirement. Also, with UG R1-only mode, adding the report req triggers Snakemake trying to run R1, R2, pair before summarizing the final report. Alternatively we just remove it altogether?
From 100 million UG 10x reads on cluster mode