Tool:gogadget: an R package for go analysis visualization and interpretation
2
10
Entering edit mode
4.7 years ago
Benn 8.1k

gogadget: an R package for go analysis visualization and interpretation

I have developed an R package, which can help in the interpretation and visualization of GO enrichment analysis. The strategy of this package is to first do goseq analysis for your RNA-seq data. Subsequently, gogadget offers functions to make sense out of these goseq results.

First function gogadget.p.adjust adds an adjusted p-value to the results.

Second function gogadget.explore explores possible (ideal) filtering settings. The idea is to filter out GO terms that are too general or too specific.

Third function gogadget.filter filters with the ideal settings (found with the previous function).

Fourth function gogadget.heatmap groups the GO terms that are similar together. Similarity is based on overlapping genes between the terms.

Fifth and sixth functions, gogadget.cytoscape and gogadget.gmt, respectively, can export your filtered results to cytoscape EnrichmentMap.

The package is developed in R 3.3.2. on windows 7, and was also tested on Ubuntu 14.04. To install the package, download the tar.gz file from https://sourceforge.net/projects/gogadget/ and use the following code in R:

install.packages( "C:/your/directory/gogadget_2.0.tar.gz", repos = NULL, type="source")


I have written an extensive user's guide, which is also available on https://sourceforge.net/projects/gogadget/. Furthermore, there is a case study which you can try (files also on sourceForge).

Please cite the following paper if you publish results obtained using gogadget:

Nota, B. (2017), Gogadget: an R Package for Interpretation and Visualization of GO Enrichment Results. Mol. Inf. May;36(5-6). doi: 10.1002/minf.201600132

Let me know if there are questions or comments. Thank you.

Ben

RNA-Seq R goseq gogadget Tool • 5.0k views
1
Entering edit mode
3.5 years ago
Benn 8.1k

NEWS 2017/10/27

I recently found out that the gogadget.gmt function didn't work properly anymore. Probably the behavior of dplyr functions changed, when I (accidentally) updated some packages.

However, this function is fixed in the newest version of gogadget v2.1.

0
Entering edit mode

Hi Ben! I'm excited to try gogadget. However, I obtained my GO enrichment results using a different tool from GOseq. Is there a way for me to format my data so that I can still use gogadget despite not using GOseq? I have the GO IDs, adjusted p-values, gene ratios, etc, in a dataframe.

Thank you, Jenny

0
Entering edit mode

Hi Jenny,

I have not tried to use a different enrichment method, but depending on what function you want to use, I think it is possible.

You'll need to rename your columns probably, so that the function will recognize what column to use. If you type in the function in your console you can see how the function works, you'll recognize which column name it will take. Good luck.

0
Entering edit mode

Hi Ben,

I also went through the gogadget user manual. I see that the heatmap has the enriched GO terms on both axes. Is there a way to generate this heatmap to compare GO terms that are enriched in different samples (GO terms on one axis, samples on the other)?

Thank you

0
Entering edit mode

Hi Jenny,

You are right, the heatmap and overlap function are to pair wise compare the GO terms from the same experiment. The idea behind it is that many GO terms are actually very similar (share the same genes) but are called slightly different. These heatmaps are meant to group the redundant GO terms.

What you want to do is more something you'll have to do with cluster profiler. Or you can try to write some code yourself, if you like coding.

0
Entering edit mode
4.2 years ago
Benn 8.1k

NEWS 2017/01/17

gogadget version 2.0 is now available. It contains a new function called gogadget.overlap. This function calculates the overlap index between two GO terms and visualizes it in a heatmap.

Furthermore the user's guide is updated, with a detailed explanation of the new function, and an extended analysis of the case study (showing how to do analysis with unfiltered GO terms).

0
Entering edit mode

Hi Benjamin.

I had downloaded one of the first versions of your software, just when it was announced. I was curious about how could I use the gogadget package with my own data, relating a non-model organism. I remember that I saw in the documentation some comments about how to work with this kind of data. But I was in a hurry at that time, and I couldn't give a try to your software.

Now I am trying gogadget again, with the new version 2.0, but I am not able to find any guidance on how to deal with non-model organism data. I even searched the version 1.0 manual, but I found nothing.

Were the tips removed from the manual? Can you tell me how could I use gogadget with my own data?

-- David da Silva Pires

0
Entering edit mode

Hi David,

Thanks for your interest, I hope you can use it for your own non-model organism.

What you need is the gene length data, and category mapping. You can find instructions in the goseq user's guide page 3 http://bioconductor.org/packages/release/bioc/vignettes/goseq/inst/doc/goseq.pdf, if that works you are probably able to further use gogadget (I have never used a non-model organism in gogadget before, so please let me know if it works eventually for you).

Good luck!

0
Entering edit mode

Hi David, I haven't heard back, so I hope no news is good news? If you're still having problem using gogadget with non-model organisms, please open a new question with some of your code in it as well, and tag it with gogadget. I can help you then with the code adjustment which is probably necessary. Some of the functions need to be altered a bit, so I can show you how to do this then. To do this here in the comments is not convenient.