How to: Use aroma.* data within Bioconductor
Author: Henrik Bengtsson
Both probe-level data and probe-summary data can be exported by aroma.affymetrix for further analysis in Bioconductor. Note that most of these methods will export data as in-memory object, because that is how most of Bioconductor is designed. For large probe-level data set this often means that it is not possible not export the complete data set. For probe-summarized data, an order of magnitude larger data sets can be exported, but for really large data sets working with in-memory object will also become a problem.
Exporting probe-summary data
Probe-summarized data can be exported in various ways depending on chip type.
ChipEffectSet as exprSet
Chip effects can be exported as an exprSet object (defined in the Biobase package) by calling:
es <- extractExprSet(ces)
where ces
is a ChipEffectSet. The extracted object will be annotated
with as much information as possible at the time of the call. Note that
this will load all data in to memory.
ChipEffectSet as eSet
Chip effects can be exported as an eSet object (defined in the Biobase package) by calling:
es <- extractESet(ces)
The extracted object will be annotated with as much information as possible at the time of the call. Note that this will load all data in to memory.
Fitted ProbeLevelModel as PLMset
Fitted ProbeLevelModel:s can be exported as an PLMset object (defined in the affyPLM package) by calling:
ps <- extractPLMset(plm)
This call requires that the PLM has been fitted for all units, otherwise
part of the results will be zero or NA
. The extracted object will be
annotated with as much information as possible at the time of the call.
Note that this will load all data in to memory.