diffbind update

Hello all,

I was wondering if there are any plans to update diffbind on galaxy.org? The current version is 3.6.1 but galaxy is currently running 2.10.0

Thanks

Hi @stealsh

The version at the top of the tool form in Galaxy is the wrapper around the 3rd party tool. If you run a tool in Galaxy, the version of the different components (requirements) actually used for the job are included in the Job Logs. Those requirements are also included in the Toolshed.

For this tool, most seems to be current between the Diffbind Bioconductor version and the Galaxy version in the ToolShed, and as hosted for use at the public servers UseGalaxy.org, UseGalaxy.eu, and UseGalaxy.org.au.

Bioconductor: “Bioconductor version: Release (3.15)”

Galaxy ToolShed: Galaxy | Tool Shed

Requirements:

name version type
bioconductor-diffbind 2.10.0 package
r-base 3.5.1 package
r-getopt 1.20.3 package
r-rjson 0.2.20 package

You mentioned version “3.6.1” – and that is Bioconductor version for their package when installed locally. Galaxy uses portions of that in the wrapper, and incorporates other parts directly.

Is there a specific sub-function when using Diffbind that you are looking for? It may not have been included in the current Galaxy wrapper that you see represented on the tool page. Or, maybe we can help you to find it :slight_smile:

Hi @jennaj,

Thanks for the detailed reply. I appreciate all the hard work you all do to make galaxy such a great resource!

There seems to be many functions in diffbind (as per the vignette) that are not available in galaxy.

For example, it is possible to output Venn diagrams in diffbind to show overlap, and lack of, between samples; this is not present in galaxy. Also, profile plots are not available in galaxy. Further, different normalisation methods are available in diffbind, such as sequencing depth or spike-in DNA. These aren’t available in galaxy, but I would like to be able to use both methods, though I assume sequencing depth is default in galaxy.

Also, this may be a separate question that I need to start a new thread for, but is it possible to download the data from biffbind separated out into different groups? What I mean by this is that I have data from two transcriptions factors, TF1 & TF2. Both bind at the same sites throughout the genome, but diffbind shows that in some loci only TF1 binds and in other loci only TF2 binds. What I’d like to do is download the coordinates for where TF1 only binds, where TF2 only binds, and where TF1&TF2 bind together as separate groups so I can do motif analysis. Is this possible with diffbind or do I need to find another way?

Regards,
Steve

1 Like

Hi @stealsh

Scroll down in the tool form and expand the section “Output Options”. I think that most of of the extra outputs you are looking for are included.

Plus you can output a copy of all the Rdata tables and the Rscript used during the processing in Galaxy if you want to parse the results in custom ways, or using methods in the Bioconductor vignette directly.

Meaning, the Rscript and Rdata tables can be used with R directly inside of Galaxy (inside interactive environments), or downloaded to use wherever you prefer to use R. Find Galaxy Interactive Environments (GIEs) in the tool panel under the section Interactive Environments. The UseGalaxy.eu server hosts the largest variety of GIEs.

One example is Rstudio and the GTN has tutorials that cover basics (moving data from a history into an environment, then back out into a history), others are used in scientific tutorials, and some cover coding languages themselves.

The basic get data in/out methods are mostly the same between the different GIEs – the tool form that launches the environment will have the function details. And some of the GIEs combine multiple programming environments (R, python, bash). The Interactive Jupyter Notebook one example, and also has a GTN tutorial.

GTN Tutorials can be searched with keywords or navigated directly.

Example GTN searches:

“rstudio” == Search Tutorials
“interactive-tools” == Search Tutorials
“jupyter” == Search Tutorials

Example topic navigation:

These are coding topics that could be applied when using IEs in Galaxy, or other places. R is included. News post explaining the goals: New Topic: Data Science Survival Kit

https://training.galaxyproject.org/ >> “Data Science Survival Kit” >> Galaxy Training!

Small warning: GIEs do not automatically track or save work into a Galaxy history the same as regular tools. That means you need to specifically track the tools or functions or datasets within the environment. If you quit out of the session, the content is lost. If the server has a hiccup, the content is lost. So be sure to save your important analysis artifacts back into the history – periodically while working then once done. That can include the entire environment plus scripts and files (all will show up as “datasets” in the history).

GIEs can be left active themselves and relaunched for short time periods (find any still active in your account under User > Active InteractiveTools, but none are quite perfect. You might need to kill a prior GIE to start a new one: do this by purging the history dataset that represents it. And, definitely save back anything you would miss to the history the environment was launched from. I know I said “save” twice already … but learning the hard way isn’t so fun :slight_smile: and everyone is very used to Galaxy automatically capturing and saving all work by default. GIEs are different.

Hope that helps more! If you find a specific function that you can’t do – or that seems to be something so commonly included that it would be helpful to add more options into the tool form GUI directly, please write back and we can make a request to the developers. They can determine what is possible or practical.

Hi @jennaj,
Thanks for the detailed reply.

I’ll look into using R to help me unlock some of these other functions/outputs.

Regards,
Steve