two species in sample with garnett

30 views
Skip to first unread message

James Hong

unread,
Mar 8, 2020, 8:30:24 PM3/8/20
to garnett-users
Hi everyone,

I have a complex sample that has both rat and human cells.

So far I've been doing pretty well using monocle3.

The issue is in the cell_data_set and the marker_file, my cell-specific genes are differentiated by hg38__genesymbol (human) and rnor6_genesymbol (rat)

As you might expect, training of the classifier gives an error saying "Not enough training samples for any cell types at root of cell type hierarchy!" but when I do a marker_check this is due to an issue with gene id conversion.

How do I get around this? Would I need to do a find & replace of the species prefixes in the cell_data_set and the marker file?

Thank you

Hannah Pliner

unread,
Mar 9, 2020, 9:40:21 AM3/9/20
to garnett-users
Hi,

I think your best bet here is to use the db="none" option and then just match the gene symbols in your marker file with your gene symbols in your cell_data_set. See here: https://cole-trapnell-lab.github.io/garnett/docs_m3/#my-species-doesn-t-have-an-annotationdbi-class-database

Please note that this removes the gene conversion parts of Garnett - so you'll need to add the hg38_ and rnor6_ prefixes to your marker file so it matches the cell_data_set.

Best,
Hannah

James Hong

unread,
Mar 9, 2020, 10:44:51 AM3/9/20
to Hannah Pliner, garnett-users
Thank you!

It’s already set up like that, so this is perfect. Will test it.

James 

Sent from my iPhone

On Mar 9, 2020, at 9:40 AM, Hannah Pliner <hpl...@gmail.com> wrote:


--
You received this message because you are subscribed to the Google Groups "garnett-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to garnett-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/garnett-users/23597a54-1117-4c54-a6f1-59dd3842489f%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages