If you have any questions, please contact our Affymetrix team at all@arabidopsis.info.
Ara-Gene-ST chips with R/Bioconductor/oligo()
We recommend R/Bioconductor - free analysis software - for most users on a budget.
We can help you get started with training documents and advice on request (although there is plenty on line already).
This is a brief Bioconductor oligo() run-through for users of Ara-Gene-ST chips.
- In this runthrough we assume basic Bioconductor knowledge and:
- You are using the aragene1.0 (for 1.1 use the appropriate library).
- We use NASCRaw and NASCset as 'variables', you are not obliged to use these names.
- Italics are notes.
- Bold is output/response.
- aragene11stattairgprobe_16.0.0.tar.gz from nmg-r.bioinformatics.nl.
- AraGene-1_0-st-v1.na32.tair10.probeset.csv.zip from www.affymetrix.com
===START OF RUNTHROUGH===
source("http://bioconductor.org/biocLite.R")
biocLite("pd.aragene.1.0.st")
then
library("pd.aragene.1.0.st")
Loading required package: XXX etc...
Using the R menu, make the working directory the one with cels in (all same type).
The command getwd() ..tells you where you are already...
celfiles<-list.celfiles()
NASCRaw<-read.celfiles(files=celfiles)
Platform design info loaded.
Reading in.....XXX etc...
NASCset<-rma(NASCRaw,target="probeset")
Background correcting
Normalizing
Calculating Expression
NOTE: 'probeset' (alternative = 'core') is the level of rma normalisation.
featureData(NASCset)<-getNetAffx(NASCset,"probeset")
NOT "TRANSCRIPT" - that doesn't appear to return anything from NetAFFX.
*LONG WAIT*
THEN to find the available headers:
varLabels(featureData(NASCset))
[1] "probesetid" "seqname" "strand" "start" "stop" "probecount"
[7] "transcriptclusterid" "exonid" "psrid" "geneassignment" "mrnaassignment" "crosshybtype"
[13] "numberindependentprobes" "numbercrosshybprobes" "numbernonoverlappingprobes" "level" "bounded" "noboundedevidence"
[19] "hascds" "probesettype"
..and to look at the annotation (e.g mrnas) for a header across a few probesets:
pData(featureData(NASCset))[20000:20010,]$mrnaassignment
===END OF RUNTHROUGH===
If you wanted, you could export the normalised data as a large delimited tab file and re-load into a GUI like oneChannelGUI or affylmGUI
Please be aware that these are multi-probe-per-exon chips; and as such are flexible in the way you consider them as a transcript unit (or many) per "gene".
It's all a bit cutting/bleeding edge... and we make no guarantees that this runthrough will work with your version of the packages since Bioconductor is always a
work-in-progress and is frequently updated.