ini.call {xps}R Documentation

Informative/Non-Informative Call

Description

Computes the Informative/Non-Informative Call for the exclusion of non-informative probe sets.

Usage

ini.call(xps.data,
         filename   = character(0),
         filedir    = getwd(),
         tmpdir     = "",
         weight     = 0.5,
         mu         = 0.0,
         scale      = 1.0,
         tol        = 0.00001,
         cyc        = 100,
         alpha1     = 0.4,
         alpha2     = 0.6,
         version    = "1.3.1",
         option     = "transcript",
         exonlevel  = "",
         xps.scheme = NULL,
         add.data   = TRUE,
         verbose    = TRUE)

xpsINICall(object, ...)

Arguments

xps.data object of class DataTreeSet.
filename file name of ROOT data file.
filedir system directory where ROOT data file should be stored.
tmpdir optional temporary directory where temporary ROOT files should be stored.
weight hyperparameter, usually set to 0.5 for version="1.3.1" and to 8.0 for version="1.3.0".
mu hyperparameter allowing to correct for potential bias.
scale scaling parameter, usually set to 1.0 for version="1.3.1" and to 2.0 for version="1.3.0".
tol termination tolerance for EM algorithm.
cyc maximum number of cycles of EM algorithm.
alpha1 a significance threshold in (0,alpha2).
alpha2 a significance threshold in (alpha1,1.0).
version version of original farms package. Currently, version="1.3.1" and version="1.3.0" are implemented. Default is version="1.3.1".
option option determining the grouping of probes for summarization, one of ‘transcript’, ‘exon’, ‘probeset’; exon arrays only.
exonlevel exon annotation level determining which probes should be used for summarization; exon/genome arrays only.
xps.scheme optional alternative SchemeTreeSet.
add.data logical. If TRUE call data will be added to slots data and detcall.
verbose logical, if TRUE print status information.
object object of class DataTreeSet.
... the arguments described above.

Details

In contrast to mas5.call this function quantifies the signal-to-noise ratio for each probe set, as described in Talloen et al. Thus, the returned p-values and detection calls have a different meaning:

The p-value that is returned estimates the signal-to-noise ratio (SNR):
P-values (SNR) of less than 0.5 indicate that there is more signal than noise and the corresponding genes are considered to be ‘informative’ for further analysis. In contrast, values greater than 0.5 indicate ‘non-informative’ genes.

The informative call is computed by thresholding the p-value as in:

call “P” if p-value < alpha1
call “M” if alpha1 <= p-value < alpha2
call “A” if alpha2 <= p-value

Here “P” should be considered as informative “I”, “M” as marginally informative, and “A” as non-informative “NI”.

The defaults for alpha1=0.4 and alpha2=0.6 are set to allow “M” calls. In order to get the same results as package ‘farms_1.3.1’, you need to set alpha1=0.5 and alpha2=0.5.

For exon/genome arrays it is necessary to supply option and exonlevel.

Following options are valid for exon arrays only:
transcript: expression levels are computed for transcript clusters, i.e. probe sets containing the same ‘transcript_cluster_id’.
exon: expression levels are computed for exon clusters, i.e. probe sets containing the same ‘exon_id’, where each exon cluster consists of one or more probesets.
probeset: expression levels are computed for individual probe sets, i.e. for each ‘probeset_id’.

Following exonlevel annotations are valid for exon arrays:
core: probesets supported by RefSeq and full-length GenBank transcripts.
metacore: core meta-probesets.
extended: probesets with other cDNA support.
metaextended: extended meta-probesets.
full: probesets supported by gene predictions only.
metafull: full meta-probesets.
ambiguous: ambiguous probesets only.
affx: standard AFFX controls.
all: combination of above.

Following exonlevel annotations are valid for whole genome arrays:
core: probesets with category ‘unique’ and ‘mixed’.
metacore: probesets with category ‘unique’ only.
affx: standard AFFX controls.
all: combination of above.

Exon levels can also be combined, with following combinations being most useful:
exonlevel="metacore+affy": core meta-probesets plus AFFX controls
exonlevel="core+extended": probesets with cDNA support
exonlevel="core+extended+full": supported plus predicted probesets

Exon level annotations are described in the Affymetrix whitepaper ‘exon_probeset_trans_clust_whitepaper.pdf’.

In order to use an alternative SchemeTreeSet set the corresponding SchemeTreeSet xps.scheme.

xpsINICall is the DataTreeSet method called by function ini.call, containing the same parameters.

Value

A CallTreeSet

Note

Since I/NI-calls distinguish only between informative and non-informative genes, the calls are identical for all samples.

Author(s)

Christian Stratowa

References

Talloen, W., Clevert D.-A., Hochreiter, S., Amaratunga, D., Bijnens, J., Kass, S., and Gohlmann, H.W.H. (2006), I/NI-calls for the exclusion of non-informative genes: a highly effective filtering tool for microarray data. Bioinformatics 23(21):2897-2902

See Also

farms, mas5.call

Examples

## first, load ROOT scheme file and ROOT data file
scheme.test3 <- root.scheme(paste(.path.package("xps"),"schemes/SchemeTest3.root",sep="/"))
data.test3 <- root.data(scheme.test3, paste(.path.package("xps"),"rootdata/DataTest3_cel.root",sep="/"))

## I/NI call
call.ini <- ini.call(data.test3,"tmp_Test3INI",verbose=FALSE)

## get data.frames
snr.ini <- pvalData(call.ini)
inf.ini <- presCall(call.ini)
head(snr.ini)
head(inf.ini)

## plot results
if (interactive()) {
callplot(call.ini)
}

rm(scheme.test3, data.test3)
gc()

[Package xps version 1.2.10 Index]