Error when reading microarray idat files from illumina

Question

[salamandra]

Hi, I tried to read an idat file with R package illuminaio:

library(illuminaio)
idatfile <- 'C:/Users/Tania/Documents/GSE26677/GSE26677_non-normalized.txt'
idat <- readIDAT(idatfile, what='all')

and also with limma package:

library(limma)
bigxfile <- 'C:/Users/Tania/Documents/GSE26677/FILE/GPL6884_HumanWG-6_V3_0_R0_11282955_A.bgx'
idat <- read.idat(idatfile, bigxfile)

and got this error:

Error in readIDAT(idatfile, what = "all") : 
  Cannot read IDAT file. File format error. Unknown magic: ID_R

Seems something is wrong with idat file, but I tried these commands with several different idat files on GEO (GSE22247, GSE79308, GSE43336, GSE26677) and always got that error.

Does anyone know how to solve this?

The top of idat file is:

ID_REF  NSP Detection Pval  SP  Detection Pval
ILMN_1762337    75.85098    0.2555995   68.39282    0.6166008
ILMN_2055271    92.8892 0.03820817  95.76293    0.02371541

Answer 1

[h.mon]

Use the readBGX() function from the illuminaio package:

library( illuminaio )
bigxfile <- 'C:/Users/Tania/Documents/GSE26677/FILE/GPL6884_HumanWG-6_V3_0_R0_11282955_A.bgx.gz'
bgx <- readBGX( bigxfile )

Note that readBGX() can read compressed and uncompressed files.

You may also want to consider the GEOquery package, see for example Cannot read Illumina data (.bgx and .txt format) in R for microanalysis .

[salamandra]

I can read bgx, but not the txt file. Do you mean the txt file is not a idat file is that it? I tried the link you suggested. Specifically tried:

library(GEOquery)
eset = getGEO('GSE26677')[[1]]

but when doing:

exprs(eset)

gets the data already quantile normalized (I know that cause gives same values as normalized data when we click in specific samples in GEO). How do I get expressionset of raw data of that study?