I cannot read data into R (read.affy)

Question

[anm17]

Hello community

I am facing the current problem.

I installed Bioconductor in R and have downloaded Raw data of a certain Series (GEO NBCI) trough R and was able to unzip it and save it in a new folder, called data. In order to read the cel files I created in the new folder 'data' a phenodata.txt file, this was done using the terminal(OSX: ls data/*.CEL > data/phenodata.txt). The phenodata was perfectly created I just adjusted it in Excel, in order to have three columns to look like this: Name FileName Target
GSM692115 GSM692115 control GSM692116 GSM692116 rhinovirus
GSM692117 GSM692117 cigarette smoke extract

I made sure to use only tabs and not spaces, saved then the file as txt (tab-delimited text)

Now when I want read this file into R I get following Error, which is not clear to me because I just have three columns with three columns names:

celfiles <- read.affy(covdesc="phenodata.txt", path="data") Fehler in read.table(filename, sep = sep, header = header, quote = quote, : more columns than column names

(Fehler=Error (german ^^))

I appreciate any answers,

M

[Jimbou]

Your data doesn't fit into the read.table function. Perhaps it's your header making the problem. Change " " to "_" and try it again. You can also post a snippet of your data table and check the help page for the ?read.affy function.

Answer 1

[Carlo Yague]

Looks like another Excel bug to me...

Sometimes, Excel saves empty cells into .txt, creating additional tabulations that mess with R read.table. To prevent this to happen, a quick solution is to reopen the file in Excel, copy and paste your data (and no other cells) into a new clean sheet. Save that sheet into .txt again (tab delimited) and you should be free of those random annoying tabulation.

Answer 2

[anm17]

Thank you for your answers,

I don't know if I should repost this as a new question but I tried to solve the problem with creating the phenodata file in R:

pheno_data <- data.frame( c( # Name "GSM692115.CEL", "GSM692116.CEL",....), c( # FileName "GSM692115.CEL", "GSM692116.CEL", ..."), c( # Target "control", "rhinovirus", "cigarette smoke extract", "rhinovirus and cigarette smoke extract",.....)) colnames(pheno_data) <- c("Name", "FileName", "Target") write.table(pheno_data, "datafile/phenodata.txt", row.names=FALSE, quote=FALSE)

I made sure that in the folder 'datafile' is the same amount of .cel files and same filenames like in the phenodata. And i deleted all other .chp files. But still i get the same error.

Interestingly I tried the same procedure with another Series of GEO, but the raw data included just the .cel, and it did work.

So I just downloaded the .cel files from the the first Series I'm investigating, but same error: more columns than column names....

I really don't see why is it not working for my Series :/

[Carlo Yague]

Maybe you can try the fill=T parameter in the read.affy() function and see what it does.

[anm17]

Unfortunately it did not help, but thank you very much!