Combing two microarray datasets

Question

[hothriananya]

Hi

Can anyone help me in sorting the below error

Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), : 'data' must be of a vector type, was 'NULL'

I have been trying to combine two microarray datasets by using virtualArray package in R/Bioconductor but i am ending up with the error when I am trying to build a array.

[Devon Ryan]

It's usually easier if you show the command that produced the error.

[hothriananya]

> my_virtualArrays <- NULL
> my_virtualArrays$iPSC_hESC_noBatchEffect <- virtualArrayExpressionSets() #-- The error is after this command

[komal.rathi]

I think if you paste more code, from the start, and something reproducible, it will be more helpful.

[shirkionglu]

Hi Komal,

This is the examples that I followed:

library(virtualArray)
options(width=72)
library("GEOquery")
GSE23402 <- getGEO("GSE23402",GSEMatrix=T,AnnotGPL=FALSE)
GSE26428 <- getGEO("GSE26428",GSEMatrix=T,AnnotGPL=FALSE)
GSE23402 <- GSE23402[[1]][,1:24]
GSE26428 <- GSE26428[[1]]
summary(exprs(GSE23402)[,1:3])
summary(exprs(GSE26428))
exprs(GSE23402) <- log2(exprs(GSE23402))
exprs(GSE26428) <- (exprs(GSE26428)/20*16)
annotation(GSE23402) <- "hgu133plus2"
annotation(GSE26428) <- "hgug4112a"
my_virtualArrays <- NULL
options(width=60)
if(require(BiocParallel))
register(MulticoreParam(verbose=TRUE))
my_virtualArrays$iPSC_hESC_noBatchEffect <- virtualArrayExpressionSets()

Then I got this error:

> my_virtualArrays$iPSC_hESC_noBatchEffect <- virtualArrayExpressionSets()
Now preprocessing raw data of GSE23402: Loading annotations...
Usinghgu133plus2as Bioconductor annotation package for datasetGSE23402.
Loading required package: hgu133plus2.db
Loading required package: AnnotationDbi
Loading required package: GenomeInfoDb
Loading required package: org.Hs.eg.db
Loading required package: DBI


Now preprocessing raw data of GSE23402: Collapsing expression values to their median...
Using identifier as id variables
Now preprocessing raw data of GSE23402: Annotating expression values with SYMBOL...
Now preprocessing raw data of GSE26428: Loading annotations...
Usinghgug4112aas Bioconductor annotation package for datasetGSE26428.
Loading required package: hgug4112a.db

Now preprocessing raw data of GSE26428: Collapsing expression values to their median...
Using identifier as id variables
Now preprocessing raw data of GSE26428: Annotating expression values with SYMBOL...
Error: 2 errors; first error:
  Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), : 'data' must be of a vector type, was 'NULL'

For more information, use bplasterror(). To resume
  calculation, re-call the function and set the
  argument 'BPRESUME' to TRUE or wrap the previous
  call in bpresume().

First traceback:
  22: virtualArrayExpressionSets()
  21: lapply(all_expression_sets, virtualArrayBuildfData, collapse_fun = collapse_fun, 
          identifier = identifier)
  20: lapply(all_expression_sets, virtualArrayBuildfData, collapse_fun = collapse_fun, 
          identifier = identifier)
  19: bplapply(X, FUN, ..., BPRESUME = BPRESUME, BPPARAM = BPPARAM)
  18: bplapply(X, FUN, ..., BPRESUME = BPRESUME, BPPARAM = BPPARAM)
  17: bplapply(X = X, FUN = FUN, ..., BPRESUME = BPRESUME, BPPARAM = SerialParam(catch.errors = BPPARAM$catch.errors))
  16: bplapply(X = X, FUN = FUN, ..., BPRESUME = BPRESUME, BPPARAM = SerialParam(catch.errors = BPPARAM$catch.errors))

Please help

[komal.rathi]

I am getting exactly the same error, and I referred to the manual step by step. I am still trying to figure out what's wrong but I am pretty sure you are not doing anything incorrect, it is something with the code.

[shirkionglu]

Hi Komal,

do you think there's any ways of getting round it?

thanks

[Devon Ryan]

It's probably simpler to simply post this to the bioconductor email list and CC the maintainer (aheider at trm.uni-leipzig.de). This looks like either a bug in the package or an error in the manual. Either way, it'd be simpler to just have the maintainer fix that.

[Devon Ryan]

Hopefully you get a speedy reply from the maintainer (I just saw that you posted to the bioconductor list) :)

[shirkionglu]

fingers crossed

thanks

[shirkionglu]

hi,

I have also been trying out this package and been getting the same error as hothriananya. Anyone could kindly advise please?

[liz87.s]

Hi Sirus,

I have the same problems with this packages! and after adding your modified code in the virtualArrayExpressionSets function, there is another error for the function calculateOverlaps; unfourtunately the link that you have posted does not work :(. Could you help me with this function which R does not find? thks

Answer 1

[Sirus]

It seems there is a problem with the code (maybe was not updated by the authors)

I did the following changes in the virtualArrayBuildfData function, I remove the lines 62 and 63:

rownames(exprs_z$data) <- exprs_z$labels[[1]][, 1]  #this line
colnames(exprs_z$data) <- exprs_z$labels[[2]][, 1]  #this line
exprs_z <-  exprs_z$data

And replaced it as following (highlighted lines)

message("Now preprocessing raw data of ", as.character(x), 
        ": Annotating expression values with ", as.character(identifier), 
        "...")
    #rownames(exprs_z$data) <- exprs_z$labels[[1]][, 1]  #this line
    #colnames(exprs_z$data) <- exprs_z$labels[[2]][, 1]  #this line
    rownames(exprs_z) <- exprs_z$identifier  #this line
    exprs_z <- exprs_z[,-1]  #this line
    exprs(z) <- data.matrix(exprs_z, rownames.force = TRUE)
    match_fData_y2 <- match(rownames(exprs(z)), fData(z)[, 2])
    fData(z) <- fData(z)[match_fData_y2, ]

In my case the package cannot find the function calculateOverlaps if you get the same issue you can copy past the code from here

Hope it helps (even if it is 5 months later)