dviraran
commented
4 days ago Hi,
I'm not sure what is the issue with the website. However, for all those other signatures, we just run ssGSEA with the GSVA package. You don't need the xCell package for that (and it's not supposed to work).
Best, Dvir
Hello,
I am having issues running xCell using the Charoentong signatures. This analysis won't run on the webpage and produes the error below. If I select the xCell signatures and use the same dataset the analysis runs successfully, so I don't think it's a problem with my input. "xCell results: Dear xCell user, We were unable to process the gene expression. The input data should be tab delimited text file of gene expression values. The rows must be human gene symbols. For xCell to work properly the full gene expression profile with all analyzed genes must be uploaded, not a subset of the genes. Please fix the data and try again. Sincerely, xCell team"
Since I couldn't get the webpage to work, I tried to run the analysis using the R package but am running into the same issue as #75. I have tried running both rawEnrichmentAnalysis and xCellAnalysis, and both produce the following error:
rawEnrichmentAnalysis(expr = exprMatrix, signatures = cell_types, genes = rownames(exprMatrix))[1] "Num. of genes: 25103" Warning: MulticoreParam() not supported on Windows, use SnowParam()Setting parallel calculations through a MulticoreParam back-end with workers=4 and tasks=100. Estimating ssGSEA scores for 28 gene sets. [1] "Calculating ranks..." [1] "Calculating absolute values from ranks..." |===========================================================================| 100% Error in model.frame.default(formula = scores ~ cell_types) : variable lengths differ (found for 'cell_types')
If I don't supply my own signatures, the xCellAnalysis runs using default signatures just fine.
Can you please advise? Fiona