[BioC] AB Rat Genome Survey Microarray

[Iain Gallagher]

Hello List & Yongming Andrew Sun


I've been tasked with the analysis of some AB Rat Genome Survey Microarray data. The data I have are text files with the following columns:

Assay_Name Row Column Probe_ID Probe_Type Gene_ID X Y Assay_Normalized_Signal Signal CL_Sig CL_Raw SDEV CV S_N CL_Sig_Error CL_Raw_SDEV Flags Sample_Name 


I was planning to use the ABarray package for this analysis but reading through the vignette I'm not following the examples. From the vignette: 


"Data File. The rows of the file represent probes. The first column is probeID (or Probe Name), the second
column is geneID, next set of columns should contain Signal, S/N and FLAGS for all samples... If array name in the header of data file is present and
arrayName is defined in experiment design file, arrayName will be used to distinguish which column is for which
hybridization sample."


So, I take my array files and create a large text file representing all arrays (20) in long format.


> arrays <- read.table('allArrays.csv', header=T, sep='\t')
> head(arrays)
  arrayName Probe_ID    Gene_ID   Signal   S.N Flags
1 GSM517967 20693443   rCG63297   264.31 -0.88     1
2 GSM517967 20693548 rCG43972.1  3258.81 15.17     0
3 GSM517967 20693561   rCG44662  1200.68 -0.22     1
4 GSM517967 20693609   rCG59171 30762.77 40.18     0
5 GSM517967 20693611   rCG58434   838.39  2.13     0
6 GSM517967 20693655   rCG21764  5331.88 20.66     0
...
537135 GSM524823 22425691   rCG48508   535.35  1.27     0
537136 GSM524823 22425699   rCG40754   518.45 -0.26     1
537137 GSM524823 22425720   rCG31597   975.71  3.62     0
537138 GSM524823 22425784 rCG61959.1   331.31 -1.62     1
537139 GSM524823 22425829   rCG38412 27281.49 45.11     0
537140 GSM524823 22426049   rCG34097  2172.00  3.65     0


I create a design file as suggested in the vignette:

> des <- read.table('design.csv', header=T, sep='\t')

> head(des)
  sampleName arrayName phenotype
1    LV_LRT1 GSM492839       LRT
2    LV_LRT2 GSM517967       LRT
3    LV_LRT3 GSM517968       LRT
4    LV_LRT4 GSM517969       LRT
5    LV_LRT5 GSM517970       LRT
6    LV_LRS1 GSM518311       LRS

Then, trying to create the expressionSet object:

> eset <- ABarray('allArrays.csv', 'design.csv', 'phenotype')
Using arrayName to match experiment with signal in file: allArrays.csv 
Error in `[.data.frame`(pd, , group) : undefined columns selected


Hmm, could someone let me know where I'm going wrong?

Thanks & best

Iain