dviraran / xcell Goto Github PK

Hello,

I am completely new in this topic and I am trying to understand what exactly include TCGA file from http://xcell.ucsf.edu/. For now, I know that there are enrichment scores (ES) across all TCGA samples for 64 cell types, but are this ES raw scores or adjusted? It is not clear for me how to interpret those results at all.

In the README you wrote:

the main usage is for comparing across samples, not across cell types

so can I use those scores to comparing each sample separately? I mean, eg: score for melanocytes for sample X is equal to 0.4 and for sample Y is equal to 0.3. Does it mean that in X sample melanocytes are more enriched? Why I shouldn't compare ES across cell types? If CD8+ T-cells (4.773e-02) have higher ES than CD4+ T-cells (2.485e-18), it is not mean that CD8+ T-cells within the same sample are more enriched? It is very unclear to me.

Additionally, I noticed that the range of scores across samples is different eg: sample X has from -2.048e-17 to 5.515e-01, and Y has from -1.430e-17 to 1.853e+00. How to interpret those difference? And the last question what do the negative values mean?

I will be very grateful for your response.

Best,
Kasia

Hi !

I see that it is possible to use custom gene signature in a GMT object format. I have a CSV file of the expression of signature gene sets for some cell types. I wonder if you could provide some example code to illustrate how to exactly construct such gene signature object from such an expression matrix?

Thanks!

Hi, I try to install from Github using devtools and I get these error messages:

devtools::install_github('dviraran/xCell')
Downloading GitHub repo dviraran/xCell@master
Error in utils::download.file(url, path, method = download_method(), quiet = quiet, :
cant not open URL'https://api.github.com/repos/dviraran/xCell/tarball/master'

I change the language ,but there is nothing change

sessionInfo()
R version 3.5.2 (2018-12-20)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

Sys.setlocale(category = "LC_ALL", locale = "us")

Would you help me work out how to fix the problem? thank you very much!

I'm interested in using this method - is there a Docker image for this repository? It would supremely help our analysis. If not would you be open to me creating one?

Hello,

I've run xCell on my dataset, and have been trying to input the results to xCellView. However every time the heatmap is almost loaded, I receive an error saying "Disconnected from the server". No matter how many times I try to reload the dataset I get the same error.

Is there any way to run xCellView locally? Or do you know what is causing the error?

I'd be very grateful for any advice you may be able to offer. Thanks!

Hi,
First, thank you for this tool.

I have been using xCell in a dataset which was generated with sorted CD3+ cells, so I decided to select only the relevant cell types from the xCell data. After doing this, I run xCell analysis but I got a warning saying that some of the genesets had length of 1, so I decided to investigate which ones were them (based on my list of genes and the list of genes in each of the signatures). I was surprised to see that the only genesets with only 1 gene are for epithelial cells, which I did not include in the list of cell types. Does xCell run the enrichment for every cell type despite the selection? (I understand that the spillover calculation is done only with the selected cell types).

Also, it might be useful to include in the documentation that the cell types should be selected from the list included in row.names( xCell.data$spill$K).

Many thanks!
Eduardo

I not able to download the xCell analysis results. When I click the link to the results it does not work. could you look into this?

Very interesting paper! Before I try to apply it to my own data, I tried to reproduce some of the data (the TCGA BRCA dataset) in the paper to make sure I’ am using it correctly. But the xCell normalized scores I get do not match what I see in the paper. Here are my steps:

1. Download the datasets.

For the TCGA FPKM measurements, from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM1536837 I downloaded the following file: GSM1536837_01_27_15_TCGA_20.Illumina.tumor_Rsubread_FPKM.txt.gz

For what TCGA samples in the above file were BRCA cancer subtype I downloaded GSE62944_01_27_15_TCGA_20_CancerType_Samples.txt.gz from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE62944

I download supplementary table 4 from http://biorxiv.org/content/early/2017/03/08/114165.figures-only

2. My code:

library(data.table)
library(xCell)

tcga.fpkm <- fread("GSM1536837_06_01_15_TCGA_24.tumor_Rsubread_FPKM.txt",sep="\t",header=T,check.names=F)
samples <- read.table("GSE62944_01_27_15_TCGA_20_CancerType_Samples.txt",sep="\t")
samples <- samples[samples[,2]=="BRCA",]
brca.fpkm <- data.frame(tcga.fpkm[,c("V1",as.character(samples[,1])),with=F],check.names=F)
row.names(brca.fpkm) <- brca.fpkm[,"V1"]
brca.fpkm <- brca.fpkm[,!(names(brca.fpkm) %in% c("V1"))]

r <- xCellAnalysis(brca.fpkm)

pdf("TCGA.BRCA.pdf",width=16,height=6)
par(mar=c(14,3,3,3))
boxplot(t(r),las=2,pch=".")
dev.off()


paper.data <- read.table("114165-5.txt",sep="\t",header=T,row.names=1)
paper.data.brca <- paper.data[,grep("BRCA",colnames(paper.data))]

pdf("TCGA.BRCA.paper.pdf",width=16,height=6)
par(mar=c(14,3,3,3))
boxplot(t(paper.data.brca),las=2,pch=".")
dev.off()

Attached are the two bar charts I get. Is there some normalizations step I’ am missing?

TCGA.BRCA.paper.pdf
TCGA.BRCA.pdf

Hi when I run xCellAnalysis I get this error message:

"Error in lapply(geneIds(mapped.gset.idx.list), function(x, y) na.omit(match(x,  : 
  could not find function "geneIds"
In addition: Warning message:
In .local(expr, gset.idx.list, ...) :
  3genes with constant expression values throuhgout the samples."

I tried searching for a package with geneid and found ballgown which has a function called "geneIDs" but it has capitol "D" so geneIDs instead geneIds as listed above. Can you help me with this? thanks!

Hi
In the documentation it is stated that the input of normalization required is gene expression normalized by gene length (RPKM, FPKM, TPM, RSEM).

Can 3' mRNA sequencing (QuantSeq) be used? Counts should not be affected by gene length as only the 3' region of the mRNA is sequenced.

Thanks

I am using xCell on PBMC RNAseq data

I am using counts per miilion TMM normalized, the main thing is surpringsly, at the levels of Neutrophils is very low if not, completely absent

I use your example data analyzed as microarray and i found exactly the same. Is it correct? I mean, i expected a high levels of neutrophils due to the fact that they are highly present on PBMC

Thanks in advance

Hi @dviraran,

I just stumbled upon #39 again in my current project.
Could you please create a release of the current version, s.t. the bioconda package gets updated?

Thanks,
Gregor

Hello there!

Thanks for your awesome work.
My question is about cDC annotation. You have a cell type called "cDC", which annotated as "Xonventional dendritic cells" in manuscript . However, I find little about Xonventional dendritic cells and cDC usually reffers to "conventional dendritic cells". I wonder what is the excatly meaning of cDC in xCell. Dose "Xonventional dendritic cells" equal to "conventional dendritic cells" ?
Thank you very much.

I and my colleagues have used xCell with both RNA seq.- and microarray-based gene expression data.

For microarray data, we have incorrectly used the xCellAnalysis function without setting its rnaseq parameter to false.

While this is completely our fault, we must say that we had simply followed the usage instruction at https://github.com/dviraran/xCell#usage and did not realize that xCellAnalysis assumes RNA seq.-based input data by default.

Mentioning the rnaseq parameter in the usage instruction on xCell's github page may prevent others from using xCell incorrectly.

Hi

I've question, whenever xCell performs the ssGSEA analysis, it only uses the ~10,000 genes from the immune signatures, right? So, no problem if the input matrix is a 60,000 transcript or just 20,000 protein coding genes, right?

On the other hand, when you perform the ssGSEA using GSVA, do you account for the maximum and minimum peak of the enrichment? From the manuscript I understand that the genes from signature are differentially expressed ¿both directions, up/down? How do you account for signatures having genes up and down regulated?

And finally, I've not been able to find what exactly the scale option does.

Thanks

Hi,
Thank you for sharing this great tool along with detailed guidelines for running the tool and interpreting the results!
I have a question regarding comparing the enrichment scores for a cell type across samples. If the samples fall into two groups, I wonder if it would be appropriate to apply a statistical test to the enrichment score to determine if samples in one group have significantly higher enrichment scores than those in the other group for a given cell type. I'd appreciate any of your suggestions. Thank you!

Hi I am a user of this great tool. One question is about Macrophage annotation. You have a cell type called "Macrophage", in addition to "Macrophage_M1" and "Macrophage M2", does this "Macrophage" mean M0 type?
Thank you.

Hi @dviraran,
I've run xCell on endometrial biopsy samples, including into expression matrix purified stromal and epithelial cells, and also samples from GTEx (2 samples per tissue) to rise heterogeneity. The curious thing is, that if I run the analysis without smooth muscle cell type, the scores are perfectly below 1, and when the smooth muscle cells are added, then their scores are above 1. Maybe you can explain the possible reasons why this is happening? Could the purified cells cause this problem? Or maybe there is nothing to worry about and this is normal?
All the best!

Hi, when I try to install from Github using devtools I get these error messages:

devtools::install_github("dviraran/xCell")
Error in nchar(object, type = "chars") :
invalid multibyte string, element 1

I tried other packages on github
like devtools::install_github("klutometis/roxygen") and installed successfully.
Would you help me find out how to fix the problem?

sessionInfo()
R version 3.6.0 (2019-04-26)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 17763)

Matrix products: default

locale:
[1] LC_COLLATE=Chinese (Simplified)_China.936 LC_CTYPE=Chinese (Simplified)_China.936 LC_MONETARY=Chinese (Simplified)_China.936
[4] LC_NUMERIC=C LC_TIME=Chinese (Simplified)_China.936

attached base packages:
[1] stats graphics grDevices utils datasets methods base

loaded via a namespace (and not attached):
[1] Rcpp_1.0.1 ps_1.3.0 prettyunits_1.0.2 rprojroot_1.3-2 digest_0.6.19 crayon_1.3.4 withr_2.1.2 assertthat_0.2.1 R6_2.4.0
[10] backports_1.1.4 magrittr_1.5 rlang_0.3.4 cli_1.1.0 curl_3.3 fs_1.3.1 remotes_2.0.4 testthat_2.1.1 callr_3.2.0
[19] devtools_2.0.2 desc_1.2.0 tools_3.6.0 glue_1.3.1 pkgload_1.0.2 compiler_3.6.0 processx_3.3.1 pkgbuild_1.0.3 sessioninfo_1.1.1
[28] memoise_1.1.0 usethis_1.5.0

Hi,

I've been trying to apply xCell to some data that has batch effect issue across groups of interest. I wonder if aggregated data containing those from different batches without batch correction could be used as input for xCell scoring and if downstream comparison of a cell-type score across groups could still be reliable. The manuscript touched briefly on the robustness against batch effect in the discussion. I'd appreciate your insights.

In addition, related to a previous issue (#26), I wonder if xCell could be applied to non-human species, such as rats, mice, and ferrets.

Thank you!

Hello,

Sorry about the basic question:
I tried to load the iris_u133a_expr.txt example file, but failed. Where did I go wrong?

library(xCell)
library(data.table)
iris <- fread("iris_u133a_expr.txt")
exprMatrix = read.table(iris,header=TRUE,row.names=1, as.is=TRUE)
xCellAnalysis(exprMatrix)

for exprMatrix = read.table(iris,header=TRUE,row.names=1, as.is=TRUE)
I get:

Error in read.table(iris, header = TRUE, row.names = 1, as.is = TRUE) :
'file' must be a character string or connection

Any suggestions?
Thank you!

Peter

Our group extensively uses xCell for some research projects. Thank you very much for sharing this package with community, it works smoothly and generates reliable results.
The only issue that we have is with result interpretation in some cases. For example, for macrophage signatures (M1/M2). Taking into account that macrophages are generally tissue resident and not part of the blood stream, it is essential for us to know how the samples used for signature generation were produced. That’s why we would truly appreciate if you can clarify the origin of M2 signatures in your xCell package.

You are claiming that M2 signature consists of six subsignatures: three of those were from BLUEPRINT and three from HPCA (Mabbott et al, 2013).
In BLUEPRINT macrophage samples are not labeled M2 vs M1. It would be great if you can confirm that “alternatively activated macrophage” from BLUEPRINT were selected for M2 signature development. It would be very helpful if you can provide exact ERX sample ids that were used. Some of them were isolated from venous blood, and some – from cord blood, which makes the interpretation of the deconvolution results very challenging.
In addition, there are over 90 samples marked as “macrophages” in HPCA study. How did you select M2?

Thank you in advance!

Hi,

I am using xCell with some colorectal cancer data and we were expecting to see a lot of epithelial cells but xCell did not report any. Do you have any suggestions about the input data (normalization, etc)? Since we did not get any epithelial cells I went ahead and downloaded the TCGA COAD data from xenahub but I still do not get the same scores that are reported on the additional file in the paper. Am I doing something different/wrong?

I can't get this installed. Could you please check on this? Thank you very much!

Tao

==================================================================

devtools::install_github('dviraran/xCell')
Downloading GitHub repo dviraran/xCell@master
from URL https://api.github.com/repos/dviraran/xCell/zipball/master
Installing xCell
"C:/PROGRA1/MICROS3/ROPEN1/R-341.2/bin/x64/R" --no-site-file --no-environ --no-save --no-restore --quiet CMD INSTALL
"C:/Users/tshi/AppData/Local/Temp/Rtmpi80SKC/devtools318022cc61b7/dviraran-xCell-0575cb2" --library="C:/Program Files/Microsoft/R Open/R-3.4.2/library" --install-tests

Installation failed: Command failed (65535)

sessionInfo()
R version 3.4.2 (2017-09-28)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252 LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] pracma_2.0.7 RevoUtils_10.0.6 RevoUtilsMath_10.0.1

loaded via a namespace (and not attached):
[1] httr_1.2.1 compiler_3.4.2 R6_2.2.0 tools_3.4.2 withr_1.0.2 curl_2.6 memoise_1.1.0 knitr_1.15.1 git2r_0.18.0 digest_0.6.12
[11] devtools_1.13.4 quadprog_1.5-5

Hi,

I have questions about how to interpret the results of xCell.

Regarding the ImmuneScore, what it means exactly? It's the sum or the average of the enrichment scores or the cell populations?

Does that mean if some samples of a set have a higher ImmuneScore they have more infiltrating cells than the others?

Best
Iael

Hello !
Thank you for providing the research community with this great open source tool !
I have a couple of questions.

1. I imagine the primary output of the tool is immune cell fractions. That means the sum total of 64 immune cells in one sample is 1 or close to 1 ?
2. Could you share your R code for doing some visualization of the immune cell output data ? I see some images you have in your paper - that would be a great start. Its hard to understand this data without any viz.
3. Do you typically look only at those cell types that have a p-value of less than 0.2 as you say in the paper , and discard those cell types with p-values more than 0.2 ?
4. Could you explain how to use the file with the Raw enrichment scores ? What can it be used for ?
   Thanks, Krithika

Hi @dviraran,

I would like to create a bioconda package for xCell, so that all dependencies for immunedeconv are available on bioconda.

• Is that ok for you?
• Could you please create a release for the current version that can be used for the bioconda package?

Best regards,
Gregor

Hi,

I know it might not be accurate to analyze <5000 gene by xCell. We still would like to check the cell type enrichment though. Is there any ways we can modify the code and let it run for low common gene?

Thanks so much!

Hi, when I try to install from Github using devtools I get these error messages:

Downloading GitHub repo dviraran/xCell@master
Error in untar2(tarfile, files, list, exdir, restore_times) :
incomplete block on file

-----sessionInfo-------

R version 3.5.2 (2018-12-20)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
[1] LC_COLLATE=Chinese (Simplified)_China.936
[2] LC_CTYPE=Chinese (Simplified)_China.936
[3] LC_MONETARY=Chinese (Simplified)_China.936
[4] LC_NUMERIC=C
[5] LC_TIME=Chinese (Simplified)_China.936

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] usethis_1.4.0 devtools_2.0.1

loaded via a namespace (and not attached):
[1] Rcpp_1.0.0 rstudioapi_0.9.0 magrittr_1.5 pkgload_1.0.2
[5] R6_2.4.0 rlang_0.3.1 tools_3.5.2 pkgbuild_1.0.2
[9] sessioninfo_1.1.1 cli_1.0.1 withr_2.1.2 remotes_2.0.2
[13] yaml_2.2.0 assertthat_0.2.0 digest_0.6.18 rprojroot_1.3-2
[17] crayon_1.3.4 processx_3.2.1 callr_3.1.1 fs_1.2.6
[21] ps_1.3.0 curl_3.3 testthat_2.0.1 memoise_1.1.0
[25] glue_1.3.0 compiler_3.5.2 desc_1.2.0 backports_1.1.3
[29] prettyunits_1.0.2

I've tried to process expression matrix with TPM values from kalisto but stopped with the error:

[1] "Num. of genes: 0"
Error in expr[shared.genes, ] : incorrect number of dimensions

I also tried this file with the online version and the analysis has finished succesfully.

Link to file on dropbox

I've installed xCell successfully, but when I load the library I'm getting the following error. I've tried removing and reinstalling 'htmltools'. including installing the most recent version 0.5.0, but not sure if the version is specificed within xCell that I can't change?

Here's the error:
Error: package or namespace load failed for ‘xCell’ in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]):
namespace ‘htmltools’ 0.4.0 is already loaded, but >= 0.4.0.9003 is required

And here's the install detailing:
Downloading GitHub repo dviraran/xCell@master

GSVA (NA -> 1.32.0 ) [CRAN]
GSEABase (NA -> 1.46.0 ) [CRAN]
pracma (NA -> 2.2.9 ) [CRAN]
geneplotter (NA -> 1.62.0 ) [CRAN]
shiny (1.4.0 -> 1.5.0 ) [CRAN]
shinythemes (NA -> 1.1.2 ) [CRAN]
graph (NA -> 1.62.0 ) [CRAN]
XML (3.98-1.20 -> 3.99-0.3) [CRAN]
bit64 (0.9-7 -> 4.0.5 ) [CRAN]
blob (1.2.0 -> 1.2.1 ) [CRAN]
Rcpp (1.0.4.6 -> 1.0.5 ) [CRAN]
bit (1.1-14 -> 4.0.4 ) [CRAN]
rlang (0.4.5 -> 0.4.7 ) [CRAN]
vctrs (0.2.4 -> 0.3.4 ) [CRAN]
ellipsis (0.3.0 -> 0.3.1 ) [CRAN]
glue (1.4.0 -> 1.4.2 ) [CRAN]
httpuv (1.5.2 -> 1.5.4 ) [CRAN]
jsonlite (1.6.1 -> 1.7.1 ) [CRAN]
htmltools (0.4.0 -> 0.5.0 ) [CRAN]
later (1.0.0 -> 1.1.0.1 ) [CRAN]
promises (1.1.0 -> 1.1.1 ) [CRAN]

Hi!
First of all, many thanks for working on this package!
I was trying xCell using a single custom genes signature that I have manually downloaded from GSEA as a GMT object.
I have run this using this code:

geneSets <- getGmt("/my_folder/geneset.gmt")
res_xCell<- xCellAnalysis(df, signatures = geneSets, rnaseq = F, parallel.sz = 8)

but the output shows that there is something wrong:

[1] "Num. of genes: 10781"

Estimating ssGSEA scores for 1 gene sets.

| 0%Using parallel with 8 cores
|========================================================================| 100%

Error in quadprog::solve.QP(Dmat, dvec, t(Amat), bvec, meq = meq) :
Amat and bvec are incompatible!

df object works perfectly fine when I run this code: res_xCell<- xCellAnalysis(df, rnaseq = F, parallel.sz = 12)

Would you like to point me what should I check first?

I am trying to use custom signatures. According to the README:

xCell loads the xCell.data object wich is a list with the spill over and calibration paramters, the signatures and the list of genes it uses. However, one can use different signautres and different spill over functions to perform the analysis.

Generating the signatures is just putting the genes in the proper format. However, I am not sure how to generate the spillover matrix and calibration parameters. Do I need to generate them manually? Can I do that in xCell?

Hi,

im working on cattle rnaseq data (whole blood) and wondering whether i could use the methodology for gene deconvolution...do i need to build a signature matrix bovine specific?

Thanks,
Laura

What is the output for the analysis? Are the output values the standard enrichment scores from GSEA?
Thanks!

Hi Dvir,
I am using xCell package in R for TPM data analysis of colon biopsies, and I have some questions:

1. How can I check that my data has the ~10,000 genes or so that you are using?
2. Are there ways to validate the results or refine them (you mentioned that you can filter some problematic cell types)?
3. For filtering the results before data adjustment: I removed non significant cell types. Is it necessary to remove cell types with low scores?. What cutoff should I use for filtering low scored cell types (i.e., what is a too low cell type average score?)?
   Thanks !
   Yehudit

Hi @dviraran,

I would like to use xCell for deconvoluting some blood samples with a different reference profile.
I am wondering if you are using all the 5079 genes listed in "13059_2017_1349_MOESM1_ESM.xlsx" in your tool?

Thanks alot.

Hello,

Is it possible to run xCell with a subset of your signatures (11 cell types, 121 signatures) and add 2 cell types with custom gene signatures?
I tried to copy paste a subset of your signatures in a gmt file and add a line per added custom population, but it considers each row of your signatures as different cell populations, and does not do the average of the scores. I also tried to play with the cell.types.use parameter, but I got errors. How can I manage the several xCell signatures per population + my 2 added signatures for 2 more populations?

Best regards,

Lisa

Dears, I am facing this error:

Error in lapply(geneIds(mapped.gset.idx.list), function(x, y) na.omit(match(x, :
could not find function "geneIds"

As the question shows, I got results from xcellanalysis, but how should I see which genes are enriched under each cell type?
Hope to get your reply, thank you!

Dear Dviraran,
when using the option signature in the xCellAnalysis function and provide a different cell type signature, the algorithm fails to run because of the microenvirnoment function.
I provided a possible solution as a pull request #4

Cheers

Hi,
Recently i found that some of the gene symbols used by xCell had been outdated. For example, AIM1 and ACP3, known as CRYBG1 and ACPP now in HGNC database. Because i always annotate my datasets using latest HGNC gene symbols, is it possible that these alias would affect the results what xCell was supposed to provide?
Thanks.

Hello,

I am very excited to use xCell for my research. Since I did not find a detailed user manual, I tried to reproduce the reported scores in the xCell paper for TCGA data to make sure I got it right.

So I ran xCell on FPKM, TPM, and the normalized data by UCSC Xena (e.g. PAAD IlluminaHiSeq pancan normalized). I ran it on individual cancer types and also all of the tumor types together. But I could not reproduce the scores reported in the paper (an example attached). I am using xCell_1.1.0. I tested all of the comments in the similar issue from June 2017 on BRCA data, but no luck.

Could you provide some details/code on how you ran xCell for TCGA data? any filtration before or after the analysis?

PAAD.pdf
xCell run is the data from paper, and In-house run is the result that I am getting.

Thank you for your time in advance,

Darya

I've tried to validate xCell on single cell type RNA-seq data and found something strange.
I used two datasets: E-GEOD-77808 with NK cells and E-GEOD-62241 with B cells. MCP counter perform well on them, but xCell gave another cell type as result.

I'm attaching two files (*.info) with the information about these datasets. Also I'm attaching figures with cell type results for these datasets after xCell and MCP-counter. More for xCell I have figures with clustering of cell types, I think it may be usefull for you.

https://www.dropbox.com/s/hkslhbflew6tkot/E-GEOD-62241.info?dl=0
https://www.dropbox.com/s/ow4yjfpz0d3wht0/E-GEOD-77808.info?dl=0

1. Bcell
   MCP-counter

xCell

xCell cluster

2. MCP
   MCP-counter

xCell

xCell cluster

Hello there!

Thank you for this awesome R package. I have a question regarding exporting files. I have run xCell on bulk RNA seq data. The usual result gives me a matrix of enriched cell types across samples. However, I want a matrix of enriched cell type across the genes that were used in the process. Is there a way to export the enrichment score across the used genes and not samples?

Thanks again.

Hi,

If you define novel cell types and corresponding spill matrices, and use the main script xCellAnalysis(), an error is generated ('ERROR - not all cell types listed are available'). That's because of line 60 in R/xCell.R. Regardless of what spill you specified, line 60 checks the new cell types against the internally stored spill matrices, and some will of course be missing.

Current:
Line 60: A = intersect(cell.types.use,rownames(xCell.data$spill$K))

Suggestion:
Line 60: A = intersect(cell.types.use,rownames(spill$K))

Cheers,
aeolianine

Dear Dr.dviraran,
Recently，I have tried my best effort to study the xcell methodology,unfortunately,there remains some questions bother me.I would appreciate if you have time to help me.
1.In the generate the signatures session,many cell types only have few samples(three or less),how to definite 10th, 25th, 33.3th, and 50th percentiles of low expression (Q1q), and 90th, 75th, 66.6th, and 50th quantiles of high expression (Q21-q).
2.Why to choose the minimum cell abundance is 0.8% ,instead of 0.1%，1% etc.
3. When fit a power function，why shift the scores to 0 using the minimal score and divided by
5000?
4.The TCGA results of xCell providing Additional file 6,I found some scores are negative,these results how to interpret?
I am sorry if these questions are a bit basic.
Best wishes，
huitingxiao

Hi,

I have questions about how interpret the results of xCell.

1. I don't understand this.

Regarding this issue, scaling the scores by samples is extremely dangerous and will inevitably will result in false interpretations.

To my understanding, we often use row "scale" to visulize the results in heatmap, such as the Fig.4a from the xCell paper. So I don't get what your mean here.

2. There are 5 seletions of "Choose gene signatures" of xCell web server. The paper of xCell used "xCell (N=64)", and what's the others? Which one should be used?

3. xCellView has an option "Filter weak signatures". How does this work?

I don't fully understand methods like xCell and don't know how to interpret the results. Sorry if these questions are a bit basic.

Bests,
Yiwei Niu

Dear xCell developer,

I am not sure how to solve the following simple error and run xcell on R:

I am uploading my expression matrix ("data.txt"), which has over 1,900 common gene names with the xcell gene list, and getting the following error:

library(xCell)
exprMatrix = read.table("data.txt",header=TRUE,row.names=1, as.is=TRUE)
xCellAnalysis(exprMatrix)
[1] "Num. of genes: 1994"
[1] "ERROR: not enough genes"

Thank you for looking into this,

Regards,

Loukia