[BioC] Oligo package: paCalls question

[Alison Ziesel]

Hello Jim,

Thanks so much for the speedy reply, I appreciate it. 

At this point in my workflow I haven’t yet performed RMA. That sounds like mistake #1 on my part.

I too had found that post by Dr. Stratowa but I confess I didn’t entirely understand it. I am willing to exclude absent probes at the probeset level rather than transcript level for this project. So something like:

>library(“oligo”)
>library(“pd.hugene.2.0.st”)
>allchips <- read.celfiles(“my list of cel files”)
>eset <- rma(allchips_present, target='core')
>featureData(eset) <- getNetAffx(eset, 'transcript')

>present <- paCalls(eset, “DABG”)
>index <- apply(present, 1, function(x) sum(x < 0.01) > 5)
>eset_present <- eset[index,]


If I’m understanding this correctly, the ‘index’ object should be those probes whose X (p-value?) is less than 0.01?

Thank you again for your advice!

alison

On Sep 4, 2014, at 2:05 PM, James W. MacDonald <jmacdon at uw.edu> wrote:

> HI Alison,
> 
> When you do paCalls with method = "PSDABG", you are getting P/A calls at the 'probeset' level (which in Affymetrix terms is the probe set region or PSR, which roughly corresponds to exon-level). But when you do rma with target = "core", you are summarizing data at the 'core' or transcript level (e.g., at this level all the PSRs that interrogate a given gene are summarized together).
> 
> So you cannot use oligo to generate P/A calls at the transcript level. It's my understanding that you can do this with xps, but I am unfamiliar with that package. However, a quick search of the BioC list using 'stratowa pacalls' got as the first hit this message:  https://stat.ethz.ch/pipermail/bioconductor/2014-August/060977.html
> 
> Best,
> 
> Jim
> 
> 
> On Thu, Sep 4, 2014 at 1:36 PM, Alison Ziesel <aziesel at emory.edu> wrote:
> Hello all,
> 
> I’ve got very basic familiarity with the oligo package, but I’ve ran into a bit of trouble with the paCalls function. I’d like to restrict the probes in my eSet object to only those called present. Here’s what I’ve done thus far:
> 
> >library(“oligo”)
> >library("pd.hugene.2.0.st”)
> >allchips <- read.celfiles(“my list of cel files”)
> 
> >allchipspa <-paCalls(allchips, method=“PSDABG”, verbose=TRUE)
> >present <- rownames(allchipspa)
> >allchips_present <- exprs(allchips[present,])
> 
> Error in exprs(allchips[present, ]) :
>   error in evaluating the argument 'object' in selecting a method for function 'exprs': Error in orig[[nm]][i, , ..., drop = drop] : subscript out of bounds
> 
> Clearly my last line doesn’t do what I want it to do: filter the object allchips versus the list of present probes in the object present. I suppose at this point I should also confirm that my object allchipspa is just the present probes!  What’s the correct way to perform this filter step? I’d be grateful for any insight you may have.
> 
> > sessionInfo()
> R version 3.1.1 (2014-07-10)
> Platform: x86_64-apple-darwin13.1.0 (64-bit)
> 
> locale:
> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
> 
> attached base packages:
> [1] parallel  stats     graphics  grDevices utils     datasets  methods   base
> 
> other attached packages:
>  [1] pd.hugene.2.0.st_3.8.1 RSQLite_0.11.4         DBI_0.2-7              oligo_1.28.2
>  [5] Biostrings_2.32.1      XVector_0.4.0          IRanges_1.22.10        Biobase_2.24.0
>  [9] oligoClasses_1.26.0    BiocGenerics_0.10.0
> 
> loaded via a namespace (and not attached):
>  [1] affxparser_1.36.0     affyio_1.32.0         BiocInstaller_1.14.2  bit_1.1-12
>  [5] codetools_0.2-8       ff_2.2-13             foreach_1.4.2         GenomeInfoDb_1.0.2
>  [9] GenomicRanges_1.16.4  iterators_1.0.7       preprocessCore_1.26.1 splines_3.1.1
> [13] stats4_3.1.1          tools_3.1.1           zlibbioc_1.10.0
> 
> alison
> 
> Alison Ziesel
> Department of Ophthalmology, Emory University
> aziesel at emory.edu
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
> 
> 
> 
> -- 
> James W. MacDonald, M.S.
> Biostatistician
> University of Washington
> Environmental and Occupational Health Sciences
> 4225 Roosevelt Way NE, # 100
> Seattle WA 98105-6099 


	[[alternative HTML version deleted]]