Skip to content

Barcode tagging

Analogy with other tools

1) UMI-tools

Read groupping

The --per-cell and --per-gene attributes are used to tell UMI-tools to group reads within their respective cells and genes. But what they eactly used for read groupping is needed to be specified separately via --cell-tag and --gene-tag, which use a 2-character tag (e.g., XT) found in the tag field of each read in the .bam file.

  1. For --per-gene, it must be used in combination with --gene-tag.

  2. For --per-cell, it can optionally be used in combination with --cell-tag.

--per-gene
--gene-tag

--gene-tag
--cell-tag
--umi-tag

gene

If you have used --gene-tag, then you also need to use --assigned-status-tag too, which is a bool parameter, specified in tag region as well and first introduced by FeatureCounts.

--assigned-status-tag=XS
UMI retrieval

It uses the --umi-tag attribute to extract a UMI per read, with its respective tag stored in a .bam file. For example, to extract UMIs tagged with the MB tag, it is used in combination with the --extract-umi-method attribute. 

--extract-umi-method=tag --umi-tag=MB

When this is not available, it uses --umi-separator to get UMIs from read names. For example, to extract UMIs separated with _, it does

--umi-separator=_