[BioC] problems with estimateGLMTrendedDisp()

[Gordon K Smyth]

Dear Fernando,

I can guess what the problem is and, if I'm right, it is a bug and you 
haven't done anything wrong.  The TrendedDisp function is separating your 
observations into bins based on average expression (essentially total read 
count), then passing each bin to dispCoxReid().  However dispCoxReid() by 
default ignores all transcripts with total counts less than 5.  I suspect 
that the first bin of your data is made up entirely of transcripts with 
total counts less than 5, leading dispCoxReid() to filter all the 
available observations and hence to give an error.

I will consult with the edgeR team here to get this bug fixed.  In the 
meantime, my guess is that everything will work for you perfectly if you 
remove rows with total count less than 5.  None of these transcripts can 
possibly give significant differences in any case.

   A <- rowSums(getCounts(eset_a_tmm))
   eset_a_tmm <- eset_a_tmm[A>=5,]

etc.

Best wishes
Gordon

> Date: Fri, 29 Apr 2011 11:46:05 -0500
> From: "Biase, Fernando" <biase at illinois.edu>
> To: "bioconductor at r-project.org" <bioconductor at r-project.org>
> Subject: [BioC] problems with estimateGLMTrendedDisp()
> Content-Type: text/plain
>
> Dear list members,
>
> I am having problems using the function estimateGLMTrendedDisp()  in 
> edgeR. When I run it on my dataset, it returns the error:
>
> eset_b_tmm <- estimateGLMTrendedDisp(eset_a_tmm ,  design)
>
> Error in dispCoxReid(y, design, offset = offset, ...) :
>  no data rows with required number of counts
>
> If I run the estimateGLMCommonDisp() function in my dataset it works.
>
> If I run the example from the Documentation page it works. I have read 
> the documentation but I do not find the source of my problem.
>
> My object eset_a_tmm is a "DGEList".
>
> The only difference I see between my dataset and the y from the 
> Documentation page is that my data has samples and some genes with 0 
> (zero) counts.
>
> Does anyone have a clue of what I am doing wrong?
>
> Thanks very much,
> Fernando
>
> sessionInfo()
> R version 2.13.0 (2011-04-13)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252
> [3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
> [5] LC_TIME=English_United States.1252
>
> attached base packages:
> [1] splines   stats     graphics  grDevices utils     datasets  methods   base
>
> other attached packages:
> [1] DESeq_1.4.0     locfit_1.5-6    lattice_0.19-23 akima_0.5-4     Biobase_2.12.0
> [6] edgeR_2.2.1
>
> loaded via a namespace (and not attached):
> [1] annotate_1.30.0      AnnotationDbi_1.14.0 DBI_0.2-5
> [4] genefilter_1.34.0    geneplotter_1.30.0   grid_2.13.0
> [7] limma_3.8.0          RColorBrewer_1.0-2   RSQLite_0.9-4
> [10] survival_2.36-5      tools_2.13.0         xtable_1.5-6

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}