search

dviraran / xCell

Cell types enrichment analysis
182 stars 62 forks source link

Error in processing xcell analysis #78

Closed tjer1ryom closed 10 months ago

tjer1ryom commented 11 months ago

Hi, I have been interested in for immune infiltration analysis for a while and impressed by xcell.But many RNA expression dataset i collected from GEO cannot perform xcell and it replies like

xcell<-xCellAnalysis(GSE77167,rnaseq=F)

  • [1] "Num. of genes: 0"
  • [1] "ERROR: not enough genes"
  • Error in return - 1 : non-numeric argument to binary operator

I have checked the normalization process and each dataset contains enough gene(~20000).I cannot firure out why this happened and it have bothered me some days:( I am dull, please tell me!

dviraran commented 11 months ago

[1] "Num. of genes: 0"

This means that the rownames in your data are not gene symbols. xCell assumes that you are using gene symbols and not some other gene ids.

Best, Dvir

tjer1ryom commented 11 months ago

Thank for your reply! According to your prompt, I rechecked the column of gene symbols (rownames) and confirmed that they indeed represent genes. However, it includes pseudogenes and mitochondrial genes, etc. It may be a problem. Additionally, I attempted to integrate the gene expression data into another dataset and found that it can be analyzed normally, which confuses me. The data I am analyzing has undergone log2 transformation and limma differential analysis. I am not sure if this has an impact, or if I should analyze the original data. I've uploaded the gene expression matrix file. If you could spare a moment to see where something might have gone wrong, I'd be very grateful! Yours, Tj GSE77167_norm_exp_data.txt

tjer1ryom commented 11 months ago

The same problem also occurs here, only endured background correction, rma normalizaion and gene expression index calculation of probe intensities. Really want to konw the cause and the solution, pls QAQ Sorry for my disturbance. GSE122626_norm_exp_data.txt

dviraran commented 11 months ago

For some reason you have an extra space in all the gene names. rownames(a) = gsub(' ','',rownames(a)) should fix that...

tjer1ryom commented 10 months ago

Apologize for my silence for weeks, and thank for your reply from the bottom of my heart! However, there's another issue emerged:

The same data can be analysied in the web xcell, so i wonder what's wrong and how can i figure it out. (i have updated all packages and changed network enviroment) Thanks again! Yours, Tj

dviraran commented 10 months ago

It's an issue with the matrixStats package. It made havoc throughout many packages. You need to upgrade your BioConductor and all associated with packages.

Best, Dvir

tjer1ryom commented 10 months ago

I conquer it! tooooooooooo much appreciation for you!!! Wish you a wonderful 2024! :) Yours, Tj