FORUMcross species single cells rna seq comparison using seurat
Joaquin Henriquez asked 2 months ago



Hi,
I would like to evaluate if there are commons cluster between my single cell rna data (a specie of lizard) and a mouse single cell rna data. I been able to merge both data using seurat in r. However, I need to change all the annotation of the genes of my lizard data in order to find common cluster. Can you help to do this in r?

regards,
Joaquin.

1 Answers
flefebvre Staff answered 2 months ago



Hi Joaquin, it would depend on the “Lizard”. If you are lucky, the genome is well annotated, present in the popular databases and someone somewhere already computed putative orthologs on the species, which I understand is necessary to integrate the two datasets in Seurat. It seems authors of the Seurat Nat. biotech paper did this exercise between human and mouse assuming “identical names”, which I find embarrassing:
https://github.com/satijalab/seurat/issues/597
Anyhow, the resource I am familiar with is InParanoid:
http://inparanoid.sbc.su.se/cgi-bin/index.cgi
You can have a look at the AnnotationFuncs or the various InParanoid packages in Bioconductor to accomplish this with R, but I believe not all InParanoid mappings are in Bioconductor, so you may have to use the text files from inparanoid.sbc.su.se directly.
The web also suggests getLDS( ) from BiomaRt, this might help:
http://www.biotechworld.it/bioinf/2017/01/12/retrieve-homolog-genes-using-biomart/
Maybe other people on this forum have other suggestions.