search

grisslab / scAnnotatR

Other
15 stars 2 forks source link

Not able to use cell_types = "all" #5

Closed JH1606-code closed 2 years ago

JH1606-code commented 2 years ago

Hi,

thanks for this nice tool but I have a problem to do the analysis of my data with the parameter cell_types = "all". If I use your test sample of the tutorial (https://bioconductor.org/packages/devel/bioc/vignettes/scAnnotatR/inst/doc/classifying-cells.html) with the following parameter everything went fine:

seurat.obj <- classify_cells(classify_obj = tirosh_mel80_example, assay = 'RNA', slot = 'counts', cell_types = c('B cells', 'NK', 'T cells'), path_to_models = 'default')

But if I change the cell_types to "all" then I get the following error:

seurat.obj <- classify_cells(classify_obj = tirosh_mel80_example, assay = 'RNA', slot = 'counts', cell_types = 'all', path_to_models = 'default')

Error: snapshotDate(): 2021-10-20 loading from cache Error in marker_genes(x) : try to request a slot "marker_genes" from an object of simple class ("NULL") without slots

Do you have any idea how to fix this problem?

Best

nttvy commented 2 years ago

Hi @JH1606-code

Thanks for having been interested in our tool. This problem is because a large number of features in the tirosh_mel80_example were removed, in order to have a small example dataset ship with the package. To make the vignettes work, the remained features are the ones used by those three models (B cells, T cells, and NK).

To use all the classifiers, I suggest you use another (public) dataset where there are features available for other cell-type classifiers.

Hope this helps. Bests, Vy

JH1606-code commented 2 years ago

Thanks for your fast answer. This helps a lot!