Separating human and mouse sequence reads from RNA-seq data

[lre1234]

HI all,
A collaborator is doing an experiment in which their will be a human xenograph into a mouse model, followed by RNA-seq. Most likely there will be a mixture of human and mouse transcripts in there. Has anyone done something to deal with this? Is there a way to dissociate the two sets of transcriptomes?

[GenoMax]

Take a look at bbsplit from BBMap suite.

There are other tools like XenofilteR for this application as well.

[cmbetts]

The way I handled it in the past was to build my STAR index with both genomes/annotations simultaneously making sure to rename the chromosomes such that they're unique ("human_chr1" and "mouse_chr1" instead of both "chr1"). That way the aligner puts the reads on the correct genome and reads with too much species conservation end up as multi-mapped and not counted.