[BioC] filtering for present calls

Claire Wilson ClaireWilson at PICR.man.ac.uk
Tue Jul 19 17:28:50 CEST 2005

Hi Josh,

Not sure whether anyone has responded to your mail yet, but I think an
easier way to do this would be to load your data in, normalise it then
calculate detection calls (present/absent calls) and then use these to
filter only out those probesets called present..


# read in all the cel files in the current directory
raw.data <- ReadAffy()

# normalise using gcrma
gcrma.eset <- call.exprs(raw.data, "gcrma")

# calculate detection calls
# present/absent calls are stored in the $call slot
calls.eset <- detection.p.val(raw.data)

# show the complete present/absent calls table for the expression set, 
# columns are different chips, rows are the probesets

# summarise the calls for filtering
# effectively counts how many times a particular probeset is called
calls.sum <- rowSums(calls.eset$call == "P")

# to get all probesets that are present on all chips
present.all.chips <-
names(calls.sum[calls.sum==length(colnames(gcrma.eset at exprs))])
# filter your expression levels 
# to retrieve expression levels of only those probesets called present
on all chips
filtered.set <- gcrma.eset at exprs[present.all.chips,]

# to get all probesets that are present on x chips
# make x equal to minimum number of times a probeset must be called
present.some.chips <- names(calls.sum[calls.sum==x])

hope this helps/makes sense!


> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch 
> [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of 
> Garcia, Joshua A
> Sent: 15 July 2005 16:35
> To: bioconductor at stat.math.ethz.ch
> Subject: [BioC] filtering for present calls
> I am using the gcrma package in R to analyze affymetrix 
> HGu133A chips and 
> would like to normalize only the probes called present by 
> GCOS.  I can 
> separate the probes called present from the rest by using the 
> chp file 
> exported to excel, however, I would need to recreate it as a 
> .cel file for use 
> in R.  Is it possible to do this?  Is it possible to use an 
> excel file in R, 
> rather than a .cel file?  Any suggestions?  Any help would be much 
> appreciated.
> Thanks,
> Josh
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor

This email is confidential and intended solely for the use o...{{dropped}}

More information about the Bioconductor mailing list