Diffbind3 dba.plotMA error

Question

[alwhiteh]

Hello,

I am analyzing some ATAC-seq from flies using Diffbind3.0.8 and EdgeR.

I initially ran dba.analyze() with the default peak size of 401 and was able to graph the results using dba.plotMA and dba.plotVolcano when my contrasts were evaluated using both EdgeR and DESEQ2.

After resizing the peaks to 100 (as recommended in the update), I am able to plot the results using DESEQ2 without any issues, but the EdgeR analysis seems to be causing some errors. Specifically, when I use dba.plotMA, I get the following error:

Error in plotfun(res$Conc, res$Fold, pch = 20, cex = cex, col = crukBlue,  : 
  is.finite(xlim) are not all TRUE

When trying to export the EdgeR object using the following code:

dba.analyze(Age_ATAC, method = DBA_EDGER, bRetrieveAnalysis = T)

I encounter this error:

Unable to return DEObject: design must be present, and analysis run.

Lastly, when trying to save the EdgeR output using dba.report(), I encounter the following error:

Error in .Call2("C_solve_user_SEW0", start, end, width, PACKAGE = "IRanges") : 
  In range 11691: at least two out of 'start', 'end', and 'width', must
  be supplied.

Is there something wrong with my files (that DESEQ2 automatically filters out?), or is this a bug in the Diffbind3 update?

Thanks in advance,

Alex

[Ram]

A few points to note:

1. Please make the title specific but brief - there is no need to paste the entire error message there. I've edited it this time.
2. Please use the formatting bar (especially the code option) to present your post better. You can use backticks for inline code (`text` becomes text), or select a chunk of text and use the highlighted button to format it as a code block. I've done it for you this time.
   

[alwhiteh]

thank you! sorry about that!

[Ram]

No problem, it's just for your future reference. Welcome to the forum!

[Rory Stark]

if you could send me a copy of your Age_ATAC object, I could have a look at what is going on.

I'll also need to see the output when you call sessionInfo().

-Rory

Answer 1

[Rory Stark]

I suspect that you have an older version of DiffBind. The current version is 3.0.9 -- this issue was addressed in DiffBind 3.0.6. Check your version with package.version("DiffBind"); you can update using BiocManager::update().

[alwhiteh]

Hi Rory,

Thanks for the quick response! I double checked the diffbind version, and it was indeed 3.0.8 (i've since updated to 3.0.9), but still got the errors. I found that if I created a separate sample sheet for each contrast it solved the problem!

-Alex

[francesca3]

Hi Rory, I'm having the same problems with some of my analyses. After reading here, I downloaded the latest version of DiffBind but the error still remains. I'm getting this message when I launch the command dba.report ="Error in .new_IRanges_from_start_width(start, width) : 'start' or 'width' cannot contain NAs". There is a way to understand which regions are out of bounds? Thanks Francesca

[Rory Stark]

Francesca, if you can send me a copy of/link to the DBA object on which you are getting this error when calling dba.plotMA(), I can have a look at it.

[chill0531]

Hi Rory,

I am having the same issue that Francesca had. Were you figure out the problem that leads to the error: "Error in .new_IRanges_from_start_width(start, width) : 'start' or 'width' cannot contain NAs"?

Thank you, Connor

[Rory Stark]

What version of DiffBind are you using?

If you could send me a copy of/link to your DBA object, I can see what is causing the error.