DEXSeq estimatedispersion error

[finfin]

Hi finfin, the first messages that you are getting, are they warnings or errors?
When you changed the versions, you mention it is getting stuck in estimateDispersions, what version of DEXSeq are you using?

Hi Richard, when doing this:

> a <- fitDispersionFunction()
Warning message:
'fitDispersionFunction' is deprecated.
Use 'estimateDispersions' instead.
See help("Deprecated") 
> a
[1] "'fitDispersionFunction' is deprecated.\nUse 'estimateDispersions' instead.\nSee help(\"Deprecated\")"

Your DEXSeqDataSet object seems to be substituted for a character vector and thats the reason "testForDEU" is failing, does not seem like a nice behaviour of a deprecated function... will check this!

For both, you seem to be running DEXSeq using a single core. I would recommend using multiple cores to speed up the computations (see the parameter BPPARAM of both estimateDispersions and testForDEU). A set of 2 conditions, 3 replicates each should take less than 20 minutes using 5 cores.

Alejandro

Hi Alejandro, they are warnings.

I tried a clean environment and re-installed everything, now I got the results.
I think somehow the previous environment still used a very old version.

Thank you very much.

[areyes]

Hi finfin, the first messages that you are getting, are they warnings or errors?
When you changed the versions, you mention it is getting stuck in estimateDispersions, what version of DEXSeq are you using?

Hi Richard, when doing this:

> a <- fitDispersionFunction()
Warning message:
'fitDispersionFunction' is deprecated.
Use 'estimateDispersions' instead.
See help("Deprecated") 
> a
[1] "'fitDispersionFunction' is deprecated.\nUse 'estimateDispersions' instead.\nSee help(\"Deprecated\")"

Your DEXSeqDataSet object seems to be substituted for a character vector and thats the reason "testForDEU" is failing, does not seem like a nice behaviour of a deprecated function... will check this!

For both, you seem to be running DEXSeq using a single core. I would recommend using multiple cores to speed up the computations (see the parameter BPPARAM of both estimateDispersions and testForDEU). A set of 2 conditions, 3 replicates each should take less than 20 minutes using 5 cores.

Alejandro

[Richard Finney]

Now it's giving me this ...
> ecs <- estimateSizeFactors(dxd)
> ecs <- estimateDispersions(ecs, fitType="local" )
using supplied model matrix
using supplied model matrix
> ecs <- fitDispersionFunction(ecs)
Warning message:
'fitDispersionFunction' is deprecated.
Use 'estimateDispersions' instead.
See help("Deprecated")
> ecs <- testForDEU(ecs)
Error in (function (classes, fdef, mtable) :
unable to find an inherited method for function 'sizeFactors' for signature '"character"'
Calls: testForDEU -> sizeFactors -> <Anonymous>
Execution halted



rrrrghhh

[finfin]

estimateDispersionFit doesn't appear to be part of DexSeq

I'm not sure what package it's part of.

I checked the source code, and found in R/methods.R, where defines all the functions. In function estimateDispersions.DEXSeqDataSet<-function(fitType=c("parametric", "local","mean"))
But this parameter could not be modified in the function call.
I'm trying to set it to "local" or "mean" only as you suggested. Since I am not quite familiar with R, I am still trying to re-package and install it so that I could test how it goes.

Thank you very much.

[Richard Finney]

estimateDispersionFit doesn't appear to be part of DexSeq

I'm not sure what package it's part of.

[finfin]

Yeah, my script ran for a several days before printing this message....

In estimateDispersionsFit(object, fitType = fitType, quiet = quiet) :
the parametric fit of dispersion estimates over the mean of counts
failed, which occurs when the trend is not well captured by the
function y = a/x + b. A local regression fit is automatically performed,
and the analysis can continue. You can specify fitType='local' or 'mean'
to avoid this message if re-running the same data.
When using local regression fit, the user should examine plotDispEsts(dds)
to make sure the fitted line is not sharply curving up or down based on
the position of individual points.

Did you find this function in source code?

[Richard Finney]

Yeah, my script ran for a several days before printing this message....

In estimateDispersionsFit(object, fitType = fitType, quiet = quiet) :
the parametric fit of dispersion estimates over the mean of counts
failed, which occurs when the trend is not well captured by the
function y = a/x + b. A local regression fit is automatically performed,
and the analysis can continue. You can specify fitType='local' or 'mean'
to avoid this message if re-running the same data.
When using local regression fit, the user should examine plotDispEsts(dds)
to make sure the fitted line is not sharply curving up or down based on
the position of individual points.

[finfin]

I'm going to put some print statements in the source and try and figure it out.
I don't even get a message. it just spins for days.

Last thing I see is "using supplied model matrix".

I did not get any error message before it threw out the ones above.

Did you find anything?

[Richard Finney]

I'm going to put some print statements in the source and try and figure it out.
I don't even get a message. it just spins for days.

Last thing I see is "using supplied model matrix".

[finfin]

I'm getting locked up on estimateDispersions(), too.
Just zones out, never returns.

did you provide a flatenedfile argument to an earlier call (like the
DEXSeqDataSetFromHTSeq() function ) ???

I generated the flatten exon annotation file earlier by using the python script directly.
The commands I used were:

exons<-read.table("sampleInfo",header=T,row.names=1)
exonCount<-read.HTSeqCounts(file.path(exons$countFile),exons,"hg19.dexseq.gff")
exonCount<-estimateSizeFactors(exonCount)
exonCount<-estimateDispersions(exonCount) #It stuck here

[Richard Finney]

I'm getting locked up on estimateDispersions(), too.
Just zones out, never returns.

did you provide a flatenedfile argument to an earlier call (like the
DEXSeqDataSetFromHTSeq() function ) ???

[finfin]

Added on 1st Aug: solved. Thank you everyone.

I was trying DEXSeq and I stuck at the dispersion estimation step.
I got exactly the same error as described in:

http://seqanswers.com/forums/showthread.php?t=22644

The process stuck there for hours and then threw out error messages as following:

1: In .local(object, ...) :
Exons with less than 11 counts will be discarded. For more details read the documentation, parameter minCount
2: In .local(object, ...) :
Genes with more than 70 testable exons will be kicked out of the analysis. For more details read the documentation, parameter maxExon
3: In .local(object, ...) :
Failed to set up model frames for genes ENSG (a bunch of gene names here)

I read the previous discussion where people mentioned by upgrade to Version 1.2.1 the problem will be solved. However, I downloaded the package this week so it is the newest version.

If anyone knows how to solve this problem, please kindly reply. I do appreciate it.

I have 3 replicates for both treatment and control, and the session information is attached below.



> sessionInfo()
R version 2.15.0 (2012-03-30)
Platform: x86_64-redhat-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] stats graphics grDevices utils datasets methods
[7] base

other attached packages:
[1] pasilla_0.2.13 DESeq_1.8.3 locfit_1.5-9.1
[4] DEXSeq_1.2.1 Biobase_2.16.0 BiocGenerics_0.2.0

loaded via a namespace (and not attached):
[1] annotate_1.34.1 AnnotationDbi_1.18.4
[3] biomaRt_2.12.0 DBI_0.2-7
[5] genefilter_1.38.0 geneplotter_1.34.0
[7] grid_2.15.0 hwriter_1.3
[9] IRanges_1.14.4 lattice_0.20-6
[11] plyr_1.8 RColorBrewer_1.0-5
[13] RCurl_1.91-0 RSQLite_0.11.4
[15] splines_2.15.0 statmod_1.4.20
[17] stats4_2.15.0 stringr_0.5
[19] survival_2.36-12 tools_2.15.0
[21] XML_3.93-0 xtable_1.7-3