[BioC] problems loading imagene data with limma

Ingunn Berget ingunn.berget at umb.no
Thu May 14 15:27:25 CEST 2009

Hello all

I want to load imagene data with limma and have encountered problems

I have used the following command to read the data, and get an error:

> RG <- read.imagene(files, columns = list( f="Signal Mean",b ="Background Mean"), path=rawdata)
Read header information
Read H:/Prosjekter/ExpressionSalmon/pilotstudyHanna/Rawdata/JA005_2498Cy3.txt 
Error in RG$G[, i] <- obj[, columns$f] : 
  number of items to replace is not a multiple of replacement length

files: data frame with file names for the rawdata, 10 rows and 2 columns, one column for the Cy3 and one for the Cy5 files
rawdata:  String which defines the path

I have tried different solutions that I have found in the mailing archives, but they did not help.
Finally I discovered one column in the raw data control called "control" which were empty for all rows in the file I checked. I tried to delete this column in both of the raw data files for this array, and 

files2 <- c("testCy3","testCy5")
RG <- read.imagene(files2, columns = list( f="Signal Mean",b ="Background Mean"), path=rawdata)

This worked well for the array I tried it on (testCy3 and testCy5 are the modified rawdata files)
Now I can of course do this for all arrays, but modifying 20 files is boring and in the next experiment there will more arrays.

Question 1: Anybody who knows how I can solve this problem more elegantly?
Question 2: What is this "control" column in the Imagene files?

Best regards
Ingunn Berget (Dr. Scient)
UMB, box 5003, IHA
1432 Ås

Centre for Integrative Genetics, www.cigene.no
Centre for Biospectroscopy and Data Modelling, www.specmod.org 

More information about the Bioconductor mailing list