[BioC] DEXSeq: formulas for testForDEUTRT

Narayanan, Manikandan (NIH/NIAID) [E] manikandan.narayanan at nih.gov
Mon May 20 23:23:25 CEST 2013


Try replacing your "I(exon==exonID)" with "exon" for TRT. The default formula1 for testForDEUTRT shows why this might work, and is also different from default formula1 for testForDEU. 
> args(testForDEUTRT)
function (ecs, nCores = 1, formula0 = ~sample + condition + exon,
    formula1 = ~sample + condition * exon, dispColumn = "dispersion")


The formula applies to the following data frame, and might give you some more insight. See also here if you haven't already: https://stat.ethz.ch/pipermail/bioconductor/2013-January/050253.html
> data(pasillaExons, package="pasilla")
> modelFrameForTRT(pasillaExons)
         sample condition        type sizeFactor   exon
1    treated1fb   treated single-read         NA   this
2    treated2fb   treated  paired-end         NA   this
3    treated3fb   treated  paired-end         NA   this
4  untreated1fb untreated single-read         NA   this
5  untreated2fb untreated single-read         NA   this
6  untreated3fb untreated  paired-end         NA   this
7  untreated4fb untreated  paired-end         NA   this
8    treated1fb   treated single-read         NA others
9    treated2fb   treated  paired-end         NA others
10   treated3fb   treated  paired-end         NA others
11 untreated1fb untreated single-read         NA others
12 untreated2fb untreated single-read         NA others
13 untreated3fb untreated  paired-end         NA others
14 untreated4fb untreated  paired-end         NA others



~Mani

-----Original Message-----
From: Christopher DeBoever [mailto:cdeboeve at ucsd.edu] 
Sent: Monday, May 20, 2013 3:57 PM
To: bioconductor at r-project.org
Subject: [BioC] DEXSeq: formulas for testForDEUTRT

Hi all,

I'm trying to use DEXSeq's testForDEUTRT and provide my own model formulas but I am running into an error:

Error in unique(c("AsIs", oldClass(x))) : object 'exonID' not found

I have used similar formulas for testForDEU in the past, so I'm not sure why this isn't working. Has anyone been able to provide formulas to testForDEUTRT? I've provided my commands below.

code:

library(DEXSeq)
library(multicore)

# read files
...

exonECS =
newExonCountSet(exonCountData,design,exonGeneIDs,exonExonIDs,exonExonIntervals,exonTranscripts)
exonECS <- estimateSizeFactors(exonECS)
formuladispersion = ~ sample + (exon + condition) * group exonECS <-
estimateDispersionsTRT(exonECS,formula=formuladispersion,nCores=28)
exonECS <- fitDispersionFunction(exonECS)

formula0 = ~ sample + condition * exon + group
formula1 = ~ sample + condition * exon + group * I(exon == exonID)

exonECS <- testForDEUTRT(exonECS,formula0 = formula0,formula1 =
formula1,nCores=28)




> sessionInfo()
R version 3.0.0 (2013-04-03)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
 [7] LC_PAPER=C                 LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
[1] multicore_0.1-7    DEXSeq_1.6.0       Biobase_2.20.0
BiocGenerics_0.6.0

loaded via a namespace (and not attached):
 [1] biomaRt_2.16.0       Biostrings_2.28.0    bitops_1.0-5
 [4] GenomicRanges_1.12.2 hwriter_1.3          IRanges_1.18.0
 [7] RCurl_1.95-4.1       Rsamtools_1.12.2     statmod_1.4.17
[10] stats4_3.0.0         stringr_0.6.2        XML_3.96-1.1
[13] zlibbioc_1.6.0



Thank you,

Chris DeBoever

Frazer Laboratory
Bioinformatics and Systems Biology
University of California San Diego

	[[alternative HTML version deleted]]

_______________________________________________
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



More information about the Bioconductor mailing list