[BioC] cdf vs probe package & Linux vs PC

[w.huber at dkfz-heidelberg.de]

Hi Justin,

an AffyBatch contains data on all those probes that were found in the CEL
file. The probe package contains data on all probes for which we could get
sequence information from Affymetrix. The two don't necessarily overlap
and also nowhere in the code it is assumed that they are in the same, or
in any particular order.

What I usually to is locate the row(s) that correspond to the probe(s)
that I want to look at in the probe package, then use probepackage::xy2i
to find them in the AffyBatch.

Please let me/us know if you have more specific questions or suggestions
on how to improve the user interface for accessing the data from
individual probes. (If you know enough R, feel to look at the methods
directly

> showMethods('pm', classes='AffyBatch', inc=TRUE)
> showMethods('probes', classes='AffyBatch', inc=TRUE)

to see 'below the hood'. But of course that should not be necessary for
the normal user.)

Best wishes
  Wolfgang

-------------------------------------
Wolfgang Huber
Division of Molecular Genome Analysis
German Cancer Research Center
Heidelberg, Germany
Phone: +49 6221 424709
Fax:   +49 6221 42524709
Http:  www.dkfz.de/abt0840/whuber
-------------------------------------

On Mon, 10 May 2004, Justin Borevitz wrote:

> I've noticed that the order of probes in the 2 packages does not agree.  At
> least for barley1 and ath1121501.  Also the way probes are ordered in Linux
> and Rgui (PC) does not agree.  It could be something with the alphabetizing
> of probsets names in the 2 versions.  Its possible this is true for the
> probe package coming from Affymetrix as well, which doesn't match either
> Linux or PC ordering. Lesson never assume ordering...
>
> Maybe everyone knows this already and that is the purpose of matchprobes??
> Any help with simple calls to avoid this problem are appreciated.
>
> # in Linux
> barley.object <- read.affybatch(filenames = list.celfiles()[2])
> Warning message:
> Incompatible phenoData object. Created a new one.
>  in: read.affybatch(filenames = list.celfiles()[2])
> pnL <- rownames(pm(barley.object))
> save(pnL,file = "pnL.RData",compress=T)
> ## then download from linux to PC
>
> # On PC
> barley.object <- read.affybatch(filenames = list.celfiles()[2])
> Warning message:
> Incompatible phenoData object. Created a new one.
>  in: read.affybatch(filenames = list.celfiles()[2])
> pnPC <- rownames(pm(barley.object))
>
> load("D:/barley/pnL.RData")
>
> table(pnPC == pnL)
>  FALSE   TRUE
> 172752  78685
>
>
>
> #Observation and rough fix for probe package ordering to PC ordering
>
> setwd("d:/barley")
> library(affy)
> barley.object <- read.affybatch(filenames = list.celfiles())
> probesets <- rownames(pm(barley.object))
> length(probesets)
>
> library(barley1probe)
> length(barley1probe$Probe.Set.Names)
>
> psn <- gsub("_at[0-9]","_at",probesets)
> psn <- gsub("_at[0-9]","_at",psn)
> table(psn == barley1probe$Probe.Set.Name)
> # FALSE   TRUE
> #249955   1482
>
> setwd("d:/ath1")
> ath1.obj <- read.affybatch(filenames = list.celfiles()[1])
> aprobesets <- rownames(pm(ath1.obj))
> apsn <- gsub("_at[0-9]","_at",aprobesets)
> apsn <- gsub("_at[0-9]","_at",apsn)
> library(ath1121501probe)
> table(apsn == ath1121501probe$Probe.Set.Name)
> # FALSE   TRUE
> #   439 250639
>
> Using the x and y coords I'm reordered the probe package as follows...
>
> setwd("d:/barley")
> library(affy)
> barley.object <- read.affybatch(filenames = list.celfiles())
> pm.i <- indexProbes(barley.object, which="pm") # all genes
> pm1 <- unlist(pm.i)
> pm.i.xy <- matrix(indices2xy(pm1, abatch = barley.object),nc = 2)
> length(pm1)
> dim(pm.i.xy)
> pm.i.xy <- pm.i.xy - 1 # for affy units starting at 0.
> probesets <- rownames(pm(barley.object))
> length(probesets)
> # now match with xy in barley1probe..
> cdfxy <- paste(pm.i.xy[,1],pm.i.xy[,2])
>
> library(barley1probe)
> names(barley1probe)
> probexy <- paste(barley1probe$x,barley1probe$y)
> ordcdf <- match(cdfxy,probexy)
> psn <- gsub("_at[0-9]","_at",probesets)
> psn <- gsub("_at[0-9]","_at",psn)
> table(psn == barley1probe$Probe.Set.Name)
> # FALSE   TRUE
> #250100   1337
> table(psn == barley1probe$Probe.Set.Name[ordcdf])
> #  TRUE
> #251437
>
> barley1probe <- barley1probe[ordcdf, ]
> save(barley1probe,file = "barley1probe.RData", compress=T)
>
>
>
>
> setwd("d:/ath1")
> ath1.obj <- read.affybatch(filenames = list.celfiles()[1])
> aprobesets <- rownames(pm(ath1.obj))
> apsn <- gsub("_at[0-9]","_at",aprobesets)
> apsn <- gsub("_at[0-9]","_at",apsn)
> apsn <- gsub("_at[0-9]","_at",apsn)
> pm.i <- indexProbes(ath1.obj, which="pm") # all genes
> pm1 <- unlist(pm.i)
> pm.i.xy <- matrix(indices2xy(pm1, abatch = ath1.obj),nc = 2)
> length(pm1)
> dim(pm.i.xy)
> pm.i.xy <- pm.i.xy - 1 # for affy units starting at 0.
> # now match with xy in ath1121501probe..
> cdfxy <- paste(pm.i.xy[,1],pm.i.xy[,2])
>
> library(ath1121501probe)
> probexy <- paste(ath1121501probe$x,ath1121501probe$y)
> ordcdf <- match(cdfxy,probexy)
>
> table(apsn == ath1121501probe$Probe.Set.Name)
> # FALSE   TRUE
> #   439 250639
> table(apsn == ath1121501probe$Probe.Set.Name[ordcdf])
> # TRUE
> #251078
> ath1121501probe <- ath1121501probe[ordcdf, ]
> save(ath1121501probe,file = "ath1121501probe.RData", compress=T)
>
>
>
>
> ---
> Justin Borevitz
>
> Plant Biology
> Salk Institute
> 10010 N. Torrey Pines Rd.
> La Jolla CA, 92037
> USA
> ph. 858 453-4100X1796
> fax 858 452-4315
> mailto:borevitz at salk.edu
> http://naturalvariation.org
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
>