UMI-collapsed BAM

[gevro]

Hi, Is the output BAM contain all UMI reads from each molecule, or only one UMI-collapsed read from each unique UMI molecule? If the former, is there a way to generate a "collapsed" BAM that retains only a single read from each UMI?

[cziegenhain]

The output BAM file contains all reads. There is no option to generate a file with consensus/collapsed reads.

[gevro]

Does the gene counting / expression analysis take into account UMIs I assume? i.e., it only counts each molecule once, even if > 1 read per UMI?

[cziegenhain]

You receive a count table with both read counts and unique UMI counts. https://github.com/sdparekh/zUMIs/wiki/Output#structure-of-the-output-dgecounts-object-in-project_namedgecountsrds