DEXseq estimateDispersions error

[dingdian110]

Hi there,

I have been trying to use DEXSeq for alternative splicing analysis. I'm doing analysis following example in the DEXSeq manual. Some errors I got when doing estimateDispersions are as follows:

> dxd = estimateDispersions( dxd )
using supplied model matrix
Error in LastError$store(results = results, is.error = is.error, throw.error = TRUE) :
Errors occurred during execution. First error message:
Error in fitBeta(ySEXP = counts(object), xSEXP = modelMatrix, nfSEXP = normalizationFactors, : in call to fitBeta, the following arguments contain NA: alpha_hatSEXP

For more information, use bplasterror().
To resume calculation, re-call the function and set the argument 'resume' to TRUE or wrap
the previous call in bpresume().

First traceback:
24: estimateDispersions(dxd)
23: estimateDispersions(dxd)
22: .local(object, ...)
21: bplapply(splitObject, function(x) {
estimateDispersionsGeneEst(x, maxit = maxit, quiet = quiet,
modelMatrix = modelMatrix, niter = 10)
}, BPPARAM = BPPARAM)
20: bplapply(splitObject, function(x) {
estimateDispersionsGeneEst(x, maxit = maxit, quiet = quiet,
modelMatrix = modelMatrix, niter = 10)
}, BPPARAM = BPPARAM)
19: bpmapply(FUN, X, MoreArgs = list(...), SIMPLIFY = FALSE, USE.NAMES = FALSE,
resume = resume, BPPARAM = BPPARAM)
18: bpma

Does anyone know what this error mean? and how to fix it?

thanks a lot!

[Michael Love]

hi,

Can you show the construction of the dxd object and what the column data looks like?

[shilez]

I have the same error, and the dxd object is this:
class: DEXSeqDataSet
dim: 588047 4
exptData(0):
assays(1): counts
rownames(588047): ENSG00000000003:E001 ENSG00000000003:E002 ...
ENSG00000261841:E003 ENSG00000261842:E001
rowData metadata column names(5): featureID groupID exonBaseMean
exonBaseVar transcripts
colnames: NULL
colData names(6): sample X ... exon sizeFactor

[Michael Love]

hi,

You need to provide more information for maintainers to help debug.

Show all the code you used, including the dxd object construction, print out the sessionInfo() and paste it here, and it would be useful to know the column data (at least the structure, if you can't share the actual design information):

as.data.frame(colData(dxd))

[shikha5]

I have the same error

I suspect this error may be due to no replicates.

as.data.frame(colData(dxd))
sample condition libType exon sizeFactor
1 control SMA-control paired-end this 1.118398
2 patient SMA-patient paired-end this 0.894136
3 control SMA-control paired-end others 1.118398
4 patient SMA-patient paired-end others 0.894136

Can anyone help in fixing this error?

Thanks in advance

Shikha

[Michael Love]

just to repeat, we can't help unless you provide all the information.

1) show all the code you used, including the dxd object construction.

2) print out the sessionInfo() and paste it here

3) (shikha5 has already provided this) as.data.frame(colData(dxd))

[shikha5]

countFiles = list.files("/dati/data_svashisht/CARIPLO_Project_files/TopHat2_New_alignments_28_11_2014/dexseqCounts", full.names=TRUE)
flattenedFile = "/personal/svashisht/Cariplo_data_analysis_29-05-14/DEXSeqAnalysis/Homo_sapiens.GRCh37_chr.gff"

sampleTable = data.frame(row.names = c( "control", "patient"),condition = c("SMA-control", "SMA-patient"),libType = c( "paired-end", "paired-end"))
library( "DEXSeq" )

dxd = DEXSeqDataSetFromHTSeq(countFiles,sampleData=sampleTable,design= ~ sample + exon + condition:exon,flattenedfile=flattenedFile )

colData(dxd)
DataFrame with 4 rows and 5 columns
sample condition libType exon sizeFactor
<factor> <factor> <factor> <factor> <numeric>
1 control SMA-control paired-end this 1.118398
2 patient SMA-patient paired-end this 0.894136
3 control SMA-control paired-end others 1.118398
4 patient SMA-patient paired-end others 0.894136

head( rowData(dxd), 3 )
GRanges object with 3 ranges and 5 metadata columns:
seqnames ranges strand | featureID
<Rle> <IRanges> <Rle> | <character>
ENSG00000000003:E001 X [99883667, 99884983] - | E001
ENSG00000000003:E002 X [99885756, 99885863] - | E002
ENSG00000000003:E003 X [99887482, 99887537] - | E003
groupID exonBaseMean exonBaseVar transcripts
<character> <numeric> <numeric> <list>
ENSG00000000003:E001 ENSG00000000003 1054.5 4324.5 ########
ENSG00000000003:E002 ENSG00000000003 463.0 200.0 ########
ENSG00000000003:E003 ENSG00000000003 372.0 1352.0 ########
-------
seqinfo: 265 sequences from an unspecified genome; no seqlengths
#Normalization
dxd <- estimateSizeFactors( dxd )

#Dispersion Estimation
dxd <- estimateDispersions (dxd)
using supplied model matrix
Error: 1 errors; first error:
Error in fitBetaWrapper(ySEXP = counts(object), xSEXP = modelMatrix, nfSEXP = normalizationFactors, : in call to fitBeta, the following arguments contain NA: alpha_hatSEXP

For more information, use bplasterror(). To resume calculation, re-call
the function and set the argument 'BPRESUME' to TRUE or wrap the
previous call in bpresume().

First traceback:
23: estimateDispersions(dxd)
22: estimateDispersions(dxd)
21: .local(object, ...)
20: bplapply(splitObject, function(x) {
estimateDispersionsGeneEst(x, maxit = maxit, quiet = quiet,
modelMatrix = modelMatrix, niter = 10)
}, BPPARAM = BPPARAM)
19: bplapply(splitObject, function(x) {
estimateDispersionsGeneEst(x, maxit = maxit, quiet = quiet,
modelMatrix = modelMatrix, niter = 10)
}, BPPARAM = BPPARAM)
18: mclapply(X = X, FUN = FUN, ..., mc.set.seed = BPPARAM$setSeed,
mc.silent = !BPPARAM$verbose, mc.cores = bpworkers

sessionInfo()
R version 3.1.2 (2014-10-31)
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=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats4 parallel stats graphics grDevices utils datasets
[8] methods base

other attached packages:
[1] DEXSeq_1.12.1 BiocParallel_1.0.0 DESeq2_1.6.2
[4] RcppArmadillo_0.4.500.0 Rcpp_0.11.3 GenomicRanges_1.18.3
[7] GenomeInfoDb_1.2.3 IRanges_2.0.0 S4Vectors_0.4.0
[10] Biobase_2.26.0 BiocGenerics_0.12.1

loaded via a namespace (and not attached):
[1] acepack_1.3-3.3 annotate_1.44.0 AnnotationDbi_1.28.1
[4] base64enc_0.1-2 BatchJobs_1.5 BBmisc_1.8
[7] biomaRt_2.22.0 Biostrings_2.34.0 bitops_1.0-6
[10] brew_1.0-6 checkmate_1.5.0 cluster_1.15.3
[13] codetools_0.2-9 colorspace_1.2-4 DBI_0.3.1
[16] digest_0.6.4 fail_1.2 foreach_1.4.2
[19] foreign_0.8-61 Formula_1.1-2 genefilter_1.48.1
[22] geneplotter_1.44.0 ggplot2_1.0.0 grid_3.1.2
[25] gtable_0.1.2 Hmisc_3.14-6 hwriter_1.3.2
[28] iterators_1.0.7 lattice_0.20-29 latticeExtra_0.6-26
[31] locfit_1.5-9.1 MASS_7.3-35 munsell_0.4.2
[34] nnet_7.3-8 plyr_1.8.1 proto_0.3-10
[37] RColorBrewer_1.0-5 RCurl_1.95-4.3 reshape2_1.4
[40] rpart_4.1-8 Rsamtools_1.18.2 RSQLite_1.0.0
[43] scales_0.2.4 sendmailR_1.2-1 splines_3.1.2
[46] statmod_1.4.20 stringr_0.6.2 survival_2.37-7
[49] tcltk_3.1.2 tools_3.1.2 XML_3.98-1.1
[52] xtable_1.7-4 XVector_0.6.0 zlibbioc_1.12.0

[shikha5]

Hi All,
Now I am considering the replicates too but getting the same error as mentioned in my last post on this thread.

Can anybody help?

[areyes]

Hi @shikha5,

The error seems to be that a you have no replicates in your experiment.

As for the dataset with replication, could you add what Mike asked in post 6 and the error that you are getting?

Alejandro

[shikha5]

Hi @areyes,
Thanks for your reply. You have correctly pointed out that, the error I was getting was due to the "absence of replicates" in the analysis. But, when I included 2 replicates then it was working fine.

So, DEXSeq program for estimating dispersion requires atleast 2 repicates per sample.

Shikha Vashisht