Mouse dataset for scMuffin analysis

[ylxing0]

Hi,

I have an issue running the code below with the mouse dataset.

gsc <- prepare_gsls(gs_sources = c("PNDB","CancerSEA","msigdb"),
CM_tissues=c("Brain"),
PNDB_tissues = c("Brain"),
msigdb_hs_cat_subcat = c("Mus musculus","C2","CP:REACTOME"),
scMuffinList = scML,
genes_min = 3)

Error in apply(msigdb_hs_cat_subcat, 1, function(x) paste(x, collapse = "_")) :
dim(X) must have a positive length.

Could you please help?

Thanks,
LX.

[emosca-cnr]

Dear LX,
thanks for your interest in our tool!
I think the error is caused by your definition of msigdb_hs_cat_subcat, which must be a data.frame, see ?prepare_gsls.
So:

gsc <- prepare_gsls(gs_sources = c("PNDB","CancerSEA","msigdb"),
CM_tissues=c("Brain"),
PNDB_tissues = c("Brain"),
msigdb_hs_cat_subcat = data.frame(species="Mus musculus", category="C2", subcategory="CP:REACTOME")
scMuffinList = scML,
genes_min = 3)

should do.

[ylxing0]

Thank you so much, it worked. I encountered another issue, please see the details below:

scML <- CNV_analysis(scML, na.rm=TRUE)
IMPORTANT: CNV analysis requires gene symbols as gene identifiers!
Genes-by-cells matrix size 18718 19438
Filtering... done
Genes-by-cells matrix size 5810 7703
Preprocess object to calculate CNV...
Retrieving gene locations...
Genes with known location 4
Genes in chromosomes:
1 5 11 19
1 1 1 1
Excluding chromosomes 1 5 11 19
Consider reducing wnd_size parameter.
Calculating CNV...
Chromosome NA ...
Error in ans[[i]] : subscript out of bounds

[emosca-cnr]

The issue is that genomic location was found only for four genes: "Genes with known location 4". This is because by default CNV_analysis() was intended for human.

Please see #2 to understand how to provide a custom annotation.