cibersort can not run

[liuxy9797]

This error occurs when I run it, but otherwise works fine.

res <-deconvolute(exprMatrix, "cibersort")

Running cibersort
Error in (function (sig_matrix, mixture_file, perm = 0, QN = TRUE) :
Parameter useless(absolute = FALSE, abs_method = "sig.score")

[grst]

Hi @liuxy9797,

sorry for not following up earlier. Does this issue persist?
If yes, can you please provide a minimal reproducible example with example data?

[Gorgot]

I have the same problem. My expression matrix can run quantiseq well in immunedeconv, but not cibersort. My matrix has 50000+ rows and 120+ columes, so I think it's not the single sample error.
Here's the error dialog:
Error in (function (sig_matrix, mixture_file, perm = 0, QN = TRUE) :
unused arguments (absolute = FALSE, abs_method = "sig.score")

[grst]

I'm starting to wonder if you might have a different version of the CIBERSORT.R script.

Could you send me your CIBERSORT.R script by e-mail (do NOT post it here due to the licensing)? My email can be found on my github profile.

[Gorgot]

Thank you for your reply! I did use a wrong version of CIBERSORT.R script of v1.03. Now the issue has been settled by v1.04. By the way, may I ask which version of script is required for immunedeconv?

[grst]

I wasn't even aware there were multiple versions. I have been using v1.04.
We should probably add a check for this.

[Gorgot]

Thanks for your efforts!

[grst]

closing in favor of #137

[dfeng4127]

Hi when i run :
results <- cibersort('./type_use.txt', "./exp_sp.txt")

I got this error:
Error in (function (.x, .f, ..., .progress = FALSE) :
ℹ In index: 1.
Caused by error in predict.svm():
! Model is empty!

[LorenzoMerotto]

Hello @dfeng4127

cibersort() is not a function included in immunedeconv.
Could you please try to apply the deconvolute_cibersort() function and include the error message here?