This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to combine expression values of multiple probes for one gene?

I am a bit new to R Bioconductor and microarray analysis.

I have loaded a GEO series matrix file (GSE2990) from GEO database in R Bioconductor. This dataset contain expression values of 22283 probes. I wish to get the expression values for the genes for the dataset. Since, there are multiple probes for an individual gene in many cases, I would like to know if there is a package /R code that can combine the expression values of multiple probes for the same gene. Also does oneChannel GUI has this feature? [Please note that I wish to work with a processed GEO dataset ]

bioconductor onechannelgui r
• 6,643 views • link
updated by Ram Ram • written by ayanava18 • ayanava18

What array platform is this? Typically, if it's an Illumina Bead Array then the different probes that represent the same gene, target different parts of the gene.

• 0 views • link
updated by Ram Ram • written by andrew.j.skelton73 andrew.j.skelton73

2 answers

I see that, its an affymetrix chip. Here is snippet which would calculate mean expression of all probesets mapping to same gene.

#Download and install this package.
source("http://bioconductor.org/biocLite.R")
biocLite("hgu133a.db")

#Assuming you have CEL files
aBatch = read.affybatch(filenames = "*.CEL")

#Normalizing with gcrma
gset = gcrma(aBatch)

#fetch entrez id for all probesets.
tab = select(hgu133a.db, keys = keys(hgu133a.db), columns = c("ENTREZID"))

e = exprs(gset)
#merge probes to genes (by Mean expression)
geneExpr = t(sapply(split(tab[,1], tab[,2]), function(ids){
                    colMeans(e[ids,,drop=FALSE])
                }))

P.S: It's not recommended to do this for many reasons.

• 0 views • link
updated by Ram Ram • written by poisonAlien poisonAlien

Can you explain why it's not recommended to this?

• 0 views • link
written by ATCG ATCG

because multiple probesets from a single gene could represent multiple isoforms and by merging them you're loosing this information.

• 0 views • link
written by poisonAlien poisonAlien

Take a look at the findLargest() function in the Bioconductor genefilter package.

• 0 views • link
updated by Ram Ram • written by Sean Davis Sean Davis

Hi Sean,

Can you please help me with the exact code I need to try?

I loaded the genefilter library and tried like this, but getting those warning messages

> findLargest()
Warning message:
In read.dcf(file.path(p, "DESCRIPTION"), c("Package", "Version")) :
  cannot open compressed file 'C:/Users/Ayanabha/Documents/R/win-library/3.2/survival/DESCRIPTION', probable reason 'No such file or directory'
Error in mget(gN, getAnnMap(map, data)) :
  error in evaluating the argument 'x' in selecting a method for function 'mget': Error: argument "gN" is missing, with no default
> findLargest(gN,testStat,data="hgu133plus2")
Warning message:
In read.dcf(file.path(p, "DESCRIPTION"), c("Package", "Version")) :
  cannot open compressed file 'C:/Users/Ayanabha/Documents/R/win-library/3.2/survival/DESCRIPTION', probable reason 'No such file or directory'
Error in mget(gN, getAnnMap(map, data)) :
  error in evaluating the argument 'x' in selecting a method for function 'mget': Error: object 'gN' not found
• 0 views • link
updated by Ram Ram • written by ayanava18 • ayanava18

Log in to answer this question.

More posts like this