Question

unable to find an inherited method for function counts for signature DESeqTransform

0

Entering edit mode

Casper van Mourik • 0

@casper-peters-14190

Last seen 16 months ago

Finland

I have a problem in regards to making a variance plot from our data. I tried to use a code snippet from here (Effects of transformations on the variance): https://bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html

when trying the code snippet;

# this gives log2(n + 1) ntd <- normTransform(dds) library("vsn") notAllZero <- (rowSums(counts(dds))>0) meanSdPlot(assay(ntd)[notAllZero,])

The counts function returns the error;

Error in (function (classes, fdef, mtable) : unable to find an inherited method for function ‘counts’ for signature ‘"DESeqTransform"’

This is my sessioninfo and traceback;

> sessionInfo() R version 3.4.2 (2017-09-28) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale: [1] LC_COLLATE=English_Netherlands.1252 LC_CTYPE=English_Netherlands.1252 LC_MONETARY=English_Netherlands.1252 LC_NUMERIC=C [5] LC_TIME=English_Netherlands.1252

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

other attached packages: [1] BiocInstaller_1.26.1 vsn_3.44.0 DESeq2_1.16.1 SummarizedExperiment_1.6.5 DelayedArray_0.2.7 [6] matrixStats_0.52.2 Biobase_2.36.2 GenomicRanges_1.28.6 GenomeInfoDb_1.12.3 IRanges_2.10.5 [11] S4Vectors_0.14.7 BiocGenerics_0.22.1 RColorBrewer_1.1-2

loaded via a namespace (and not attached): Error in x[["Version"]] : subscript out of bounds In addition: Warning message: In FUN(X[[i]], ...) : DESCRIPTION file of package 'backports' is missing or broken > traceback() 7: FUN(X[[i]], ...) 6: lapply(X = X, FUN = FUN, ...) 5: sapply(L[[n]], function(x) x[["Version"]]) 4: mkLabel(x, "loadedOnly") 3: print(mkLabel(x, "loadedOnly"), quote = FALSE, ...) 2: print.sessionInfo(x) 1: function (x, ...) UseMethod("print")(x)

I have no idea what I'm doing wrong.

Thanks in advance for anyone replying to this!

deseq2 counts DESeqtransform variance deviation • 7.1k views

ADD COMMENT • link updated 7.1 years ago by Michael Love 42k • written 7.1 years ago by Casper van Mourik • 0

0

Entering edit mode

I already tried validating my install with

biocValid(), my session is valid.

ADD REPLY • link 7.1 years ago Casper van Mourik • 0

score 0 · Answer 1 · 2017-10-17

0

Entering edit mode

Michael Love 42k

@mikelove

Last seen 3 hours ago

United States

What is class(dds)? The counts() method is only defined for DESeqDataSet. You may have overwritten dds with an object of a different class.

ADD COMMENT • link 7.1 years ago Michael Love 42k

0

Entering edit mode

Running the class and typeof command returns the following (my variable has a different name);

> class(cds) [1] "DESeqDataSet" attr(,"package") [1] "DESeq2" > typeof(cds) [1] "S4"

So, seeing the output, I don't think I have a wrong object that I'm putting in.

However, looking at ntd (nds in my case) it returns a DESeqtransform object.

> typeof(nds) [1] "S4" > class(nds) [1] "DESeqTransform" attr(,"package") [1] "DESeq2"

Might this have to do with the error? If so, then the example in the doc is just plain wrong.

Edit1: code runs if I directly put in my cds (dds), so normalizing it will change the class type and cause counts to return an error.

ADD REPLY • link 7.1 years ago Casper van Mourik • 0

0

Entering edit mode

In your first post you said you called counts() on dds:

notAllZero <- (rowSums(counts(dds))>0)

This code should not give an error.

Maybe you got an error calling counts() on a DESeqTransform object? These objects don't contain counts. It contains transformed counts. The way to get out the transformed counts is using assay().

The docs (workflows and vignettes) are checked nightly and would throw an error if there were a problem, and then you wouldn't see the broken version, but an older version. Generally, if you see code in an evaluated chunk in a workflow or vignette, it should run without error on your end, given the same version of packages.

ADD REPLY • link 7.1 years ago Michael Love 42k

0

Entering edit mode

Maybe you got an error calling counts() on a DESeqTransform object?

This. I didn't see I used the wrong variable there. We have a saying in dutch; "I can't see the forest through all the trees." We use it when we're being too short-sighted / get confused because of clusterfucks (like my code.)

For anybody else ever having the same problem; double check that you put the correct object in the correct function.

ADD REPLY • link 7.1 years ago Casper van Mourik • 0