[BioC] error in mas5calls() for soybean chip
Jenny Drnevich
drnevich at uiuc.edu
Wed Apr 19 19:07:13 CEST 2006
Hi Jim and Ken and the rest of the list,
Thanks for the debugging suggestions. The problem is indeed with "probe
sets" that only have one probe pair; if the PM > MM, then 'DetectionPValue'
returns 0, but it PM <= MM then 'NaN' is returned. I was able to get
'mas5calls.AffyBatch' to work by simply adding this one command before the
P/M/A calls were made:
p[sapply(p,is.na)] <- 1
I'm not sure what's the best way to call P/M/A when there is only one probe
pair in a probe set, but perhaps a warning/output could be generated along
the lines of "One or more of the probe sets has only one probe pair. Use
'oneProbePair()' to find out which ones". [code is below for a simple
function I wrote to get the probe set names]. Or would a warning cause more
problems from unnecessary panic? I was surprised that there were 92 "probe
sets" on the soybean chip with only one probe pair? I wonder how many other
chips actually have single probe pair "probe sets" - as long as PM>MM for
all the arrays, mas5calls() would work and you'd never know...
Cheers,
Jenny
oneProbePair <- function(object) {
x <- probeNames(object)
dups <- unique(x[duplicated(x)])
singles <- setdiff(unique(x),dups)
return(singles)
}
At 01:49 PM 4/18/2006, James W. MacDonald wrote:
>Hi Jenny,
>
>Jenny Drnevich wrote:
> > Hi all,
> >
> > I've run into some sort of error or bug when trying to get Affy's
> > present/marginal/absent calls for Affy's soybean array. I've traced the
> > error to this part of mas5calls.AffyBatch():
> >
> > calls <- sapply(p, function(y) {
> > if (y < alpha1) {
> >
> >
> > According to Benjamin Otto's reponse
> > (https://stat.ethz.ch/pipermail/bioconductor/2006-January/011568.html) to
> > the exact same error message for HU6800 arrays, specifying alpha1= and
> > alpha2= in the call to mas5calls() should work, but it doesn't for soybean
> > arrays (all codes & output at the bottom). I'm actually not surprised that
> > it doesn't work, because these are the defaults in the internal function
> > mas5calls.AffyBatch() that's called. Is problem this related to similar
> > problems with the qc() function from simpleaffy not working for some array
> > types? I'll try the solution suggested in
> > https://stat.ethz.ch/pipermail/bioconductor/2006-April/012655.html, but
> > mas5calls is part of the affy package, and I didn't see where functions
> > from simpleaffy were called.
>
>This problem isn't related to anything in simpleaffy. I am betting that
>there are some NAs in the p list that is returned from the call to
>'DetectionPValue' (immediately preceding the code snippet you give
>above). Without actually having any soybean arrays to play with, I
>cannot confirm this, but you could by stepping through the code using
>
>debug(mas5calls.AffyBatch)
>mas5calls(rawdata[,1:5])
>
>when you get to the point that the 'p' list is made, you can dump it
>into your .GlobalEnv using
>
>p <<- p
>
>and then quit the debugger (using Q). You could then do something like
>
>sum(sapply(p, is.na))
>
>to see how many (or if) there are any NAs in there. You could also
>figure out which probesets are returning the NAs, and take a look at
>them using the PM() accessor to see if you can tell what the problem
>might be.
>
>Any information you might be able to give could be helpful in debugging
>this error.
>
>Best,
>
>Jim
>
>
> >
> >
> > Thanks,
> > Jenny
> >
> > > rawdata
> > AffyBatch object
> > size of arrays=1164x1164 features (317565 kb)
> > cdf=Soybean (61170 affyids)
> > number of samples=30
> > number of genes=61170
> > annotation=soybean
> >
> > > PMAcalls <- mas5calls(rawdata[,1:5])
> > Getting probe level data...
> > Computing p-values
> > Making P/M/A Calls
> > Error in if (y < alpha1) { : missing value where TRUE/FALSE needed
> >
> > > PMAcalls <- mas5calls(rawdata[,1:5],alpha1=0.04,alpha2=0.06)
> > Getting probe level data...
> > Computing p-values
> > Making P/M/A Calls
> > Error in if (y < alpha1) { : missing value where TRUE/FALSE needed
> >
> >
> > > sessionInfo()
> > R version 2.2.1, 2005-12-20, i386-pc-mingw32
> >
> > attached base packages:
> > [1] "splines" "tools" "methods" "stats" "graphics" "grDevices"
> > [7] "utils" "datasets" "base"
> >
> > other attached packages:
> > soybeancdf made4 scatterplot3d ade4 affyQCReport
> > "1.10.0" "1.4.0" "0.3-24" "1.4-0" "1.8.0"
> > simpleaffy genefilter survival reposTools affyPLM
> > "2.4.2" "1.8.0" "2.20" "1.8.0" "1.6.0"
> > affydata limma gcrma matchprobes affy
> > "1.6.0" "2.4.7" "2.2.0" "1.2.1" "1.8.1"
> > Biobase RWinEdt
> > "1.8.0" "1.7-4"
> >
> >
> >
> > Jenny Drnevich, Ph.D.
> >
> > Functional Genomics Bioinformatics Specialist
> > W.M. Keck Center for Comparative and Functional Genomics
> > Roy J. Carver Biotechnology Center
> > University of Illinois, Urbana-Champaign
> >
> > 330 ERML
> > 1201 W. Gregory Dr.
> > Urbana, IL 61801
> > USA
> >
> > ph: 217-244-7355
> > fax: 217-265-5066
> > e-mail: drnevich at uiuc.edu
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at stat.math.ethz.ch
> > 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
>Affymetrix and cDNA Microarray Core
>University of Michigan Cancer Center
>1500 E. Medical Center Drive
>7410 CCGC
>Ann Arbor MI 48109
>734-647-5623
>
>
>**********************************************************
>Electronic Mail is not secure, may not be read every day, and should not
>be used for urgent or sensitive issues.
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz.ch/mailman/listinfo/bioconductor
>Search the archives:
>http://news.gmane.org/gmane.science.biology.informatics.conductor
Jenny Drnevich, Ph.D.
Functional Genomics Bioinformatics Specialist
W.M. Keck Center for Comparative and Functional Genomics
Roy J. Carver Biotechnology Center
University of Illinois, Urbana-Champaign
330 ERML
1201 W. Gregory Dr.
Urbana, IL 61801
USA
ph: 217-244-7355
fax: 217-265-5066
e-mail: drnevich at uiuc.edu
More information about the Bioconductor
mailing list