Question

How to get missing columns in DEXSeq analysis (exon usage coefficients)

0

Entering edit mode

mjrarcher ▴ 10

@mjrarcher-18313

Last seen 5.3 years ago

Dear all,

I have been running DEXSeqAlt to perform differential exon usage analysis of a large cohort. I have three conditions in my metadata table. After calling the function DEXSeqResults (dxd), the result I obtained had all columns as shown in the vignette, but my conditions and the foldchange values. Can anyone please explain why this might be?

dexseq dexseqalt • 1.0k views

ADD COMMENT • link 5.3 years ago mjrarcher ▴ 10

0

Entering edit mode

Could you send me a small reproducible example code along with a small data example (It could be, for example, the first 100 exons of your object)?

Alejandro

ADD REPLY • link 5.3 years ago Alejandro Reyes ★ 1.9k

0

Entering edit mode

Hi Alejandro, I'm not sure how to send a reproducible example. I'll look into how I can do that, but in the mean time here is the code I used and the result I got:

#library('DEXSeqAlt')
# reducedModel <- ~ sample + exon
# fullModel <- ~ sample + exon + condition:exon
# dxd <- DEXSeqDataSetFromHTSeq(countfiles, sampleData=sampleTable, design= ~ sample + exon + condition:exon, flattenedfile = gf$
# dxd <- estimateSizeFactors(dxd)
# dxd <- estimateDispersionsAlt( dxd, formula = fullModel )
# dxd <- testForDEUAlt(dxd, reducedModel=reducedModel, fullModel=fullModel, verbose=FALSE )
# dxd = estimateExonFoldChanges( dxd, fitExpToVar="condition")
# dxr1 <- DEXSeqResultsAlt(dxd)
mcols(dxr1)$description

[1] "mean of the counts across samples in each feature/exon"      
[2] "feature/exon identifier"                                     
[3] "group/gene identifier"                                       
[4] "exon dispersion estimate"                                    
[5] "P-value of likelihood ratio test"                            
[6] "BH adjusted p-values"                                        
[7] "GRanges object of the coordinates of the exon/feature"       
[8] "matrix of integer counts, of each column containing a sample"
[9] "list of transcripts overlapping with the exon

ADD REPLY • link updated 5.3 years ago by Martin Morgan 25k • written 5.3 years ago by mjrarcher ▴ 10

0

Entering edit mode

As you can see, this is different from the values in the vignette (i'm missing columns 8, 9, and 10 in the above result):

mcols(dxr1)$description
## [1] "group/gene identifier"
## [2] "feature/exon identifier"
## [3] "mean of the counts across samples in each feature/exon"
## [4] "exon dispersion estimate"
## [5] "LRT statistic: full vs reduced"
## [6] "LRT p-value: full vs reduced"
## [7] "BH adjusted p-values"
## [8] "exon usage coefficient"
## [9] "exon usage coefficient"
## [10] "relative exon usage fold change"
## [11] "GRanges object of the coordinates of the exon/feature"
## [12] "matrix of integer counts, of each column containing a sample"
## [13] "list of transcripts overlapping with the exon"

ADD REPLY • link updated 5.3 years ago by Martin Morgan 25k • written 5.3 years ago by mjrarcher ▴ 10