I am working on affy microarray CEL file: GSM4445975

I know we can extract probe level intensity by using:

but can I still access to the probe level intensity after the rma() ?

head(exprs(data.raw)[probe_table$probe_id,])

#GSM4445944_9SR20982A_A01.CEL.gz GSM4445945_9SR20982A_A02.CEL.gz GSM4445946_9SR20982A_A03.CEL.gz
6424166                             250                              99                              83
2363351                             246                              87                             135
4847126                             830                             244                             260
2851978                             273                             109                             228
5870089                             507                             162                             231
136147                              152                             152                             212
# GSM4445947_9SR20982A_A04.CEL.gz GSM4445948_9SR20982A_A05.CEL.gz GSM4445949_9SR20982A_A06.CEL.gz
6424166                             481                             711                             150
2363351                             781                            1041                             289
4847126                            2428                            3258                             783
2851978                            1158                            1402                             289
5870089                            1688                            1898                             430
136147                              350                             494                             168

data.norm = rma(data.raw,target="probeset")
norm.exp<-exprs(data.norm)[probe_table$probe_id,]

#Error in exprs(data.norm)[ACTB$probe_id, ] : subscript out of bounds

How should I extract nomalized probe intensity?

Dear Haoran,

as far as I know, RMA summarizes probes on the probe set-level, i.e. there is no normalized probe-level value after RMA, only probe sets.

Cheers,

• axel

If you just want normalized probe values, don't summarize. You can just do

fset <- read.celfiles(list.celfiles())
fset.bg <- backgroundCorrect(fset, "rma")
fset.norm <- normalize(fset.bg, "quantile")

And then you have background corrected, normalized probe values.