GoSeq error of my data

usegalaxy.org support
, ,
1

Hello, I’m beginner of this Seq analysis starting with Galaxy.
I have studied this about 2 months.

I learned rna-seq analysis using galaxy through tutorial.

Reference-based RNA-Seq data analysis (galaxyproject.org)

I followed the tutorial perfectly using data set provided by tutorial, not any error.

but I can not execute “Goseq” process with my data, before this process everything is fine.

image

I checked my two input data ( [geneid, logic], [gene id, length])
But I checked my logic data after “compute (bool<0.05)” became short line (55402 ->17081).

image
I found the reason is my adj-p value of lots of genes are NA.

I am not sure It is reason of error.
but, I downloaded my data to my computer as a excel format.
and I intentionally changed NA to False and upload this file to galaxy history again.

image

because I think the reason of Goseq error is not the same number of row between two input data.
But It is also not excuted.

Please help me.

2

I have sort of the same problem. Trying with my data pin GOSEQ, and there is always an error, no matter what variation I try. I don’t know why.

3

Hi @Brists and @Seungyeong

Error in [.default(summary(map), , 1) : incorrect number of dimensions
Calls: run_goseq … goseq → reversemapping → [ → [.table → NextMethod

The error means that the tool couldn’t construct a matrix of the two inputs correctly. Check to make sure that the identifiers in the “Differentially expressed genes file” are all contained in the “Gene lengths file” and that the identifiers in the same sort order.

All this should already be correct if the steps in either of the tutorials below were all done. If for some reason you need to do that again, the files could be joined and split up again using the original tutorial methods or using manipulations explained in the last tutorial.

Examples of accepted formats and manipulations are here. The pre-filters remove any rows with lower scores (and NA values) before the compute step: 3: RNA-seq genes to pathways and here Reference-based RNA-Seq data analysis

There should never be a need to download data, manipulate it, then load it back to Galaxy. Especially not using Excel as that can introduce many format issues including hidden characters. If you are not sure how to do text manipulations in Galaxy, see this tutorial Data Manipulation Olympics

1 Like
4

(Just adding on to this post since it showed up when I googled for the same error message, and I saw the same error for a different case.)

In my case I’d provided Entrez IDs, but misconfigured goseq to use Gene Symbols (from a previous test case). This mismatch produces the same error message as seen in this post and it’s very unclear why. But matching those up again resolved it!

1 Like