Alejandro Reyes | 9 Oct 10:37 2012
Picon

Re: DEXSeq offset term

Dear Alicia,

Thanks for your input!

> Hi all,
>
> I am interested in using DEXSeq to look for differential expression across seven conditions, each with ~7
biological replicates. I have created an ExonCountSet object and am trying to estimate the dispersions
from my model. I have two questions: 1) Is it possible to supply an offset term to the glm? This would be very
helpful when trying to incorporate normalization info, since the count data is restricted to integers. I
noticed that the glmnb.fit object called by DEXSeq contains a mf$offset variable. Additionally, 2) Has
anyone seen and/or corrected the following error when trying to use multiple cores to estimate dispersions?

DEXSeq is designed to test for differences in exon usage, not 
differential expression.  To test for differential expression you can 
use other packages like DESeq or edgeR. Anyway, in both DESeq and DEXSeq 
you can assign normalization factors using:

sizeFactors(yourobject) <- your normalization factors

or why dont you use the normalization factors from DESeq/DEXSeq?

> ecs <- estimateDispersions(ecs, nCores=3)
> Estimating Cox-Reid exon dispersion estimates using 3 cores. (Progress report: one dot per 100 genes)
> The process has forked and you cannot use this CoreFoundation functionality safely. You MUST exec().
> Break on
__THE_PROCESS_HAS_FORKED_AND_YOU_CANNOT_USE_THIS_COREFOUNDATION_FUNCTIONALITY___YOU_MUST_EXEC__()
to debug.
> This message repeats several times and also contains the following: Tcl_ServiceModeHook: Notifier not initialized.

If you are testing for differential exon usage, do you get the error 
message using only a single core? I believe that it is a problem with 
your multicore + apple... some "googling" took me to this:

https://stat.ethz.ch/pipermail/r-sig-mac/2009-August/006426.html

Best wishes,
Alejandro

> Thanks a bunch for any potential insight!
>
>   -- output of sessionInfo():
>
>> sessionInfo()
> R version 2.15.1 (2012-06-22)
> Platform: x86_64-apple-darwin12.0.0 (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] splines   stats     graphics  grDevices utils     datasets  methods
> [8] base
>
> other attached packages:
>   [1] DEXSeq_1.2.1          Biobase_2.16.0        BiocGenerics_0.2.0
>   [4] RColorBrewer_1.0-5    scales_0.2.2          cqn_1.2.0
>   [7] quantreg_4.91         SparseM_0.96          preprocessCore_1.18.0
> [10] nor1mix_1.1-3         mclust_4.0            plyr_1.7.1
> [13] multicore_0.1-7
>
> loaded via a namespace (and not attached):
>   [1] biomaRt_2.12.0   colorspace_1.1-1 dichromat_1.2-4  hwriter_1.3
>   [5] labeling_0.1     munsell_0.4      RCurl_1.95-0     statmod_1.4.16
>   [9] stringr_0.6.1    tcltk_2.15.1     tools_2.15.1     XML_3.95-0
>
> --
> Sent via the guest posting facility at bioconductor.org.
>

_______________________________________________
Bioconductor mailing list
Bioconductor@...
https://stat.ethz.ch/mailman/listinfo/bioconductor
Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor


Gmane