Trinotate SQLite boilerplate unavailable on usegalaxy.eu?

Hi All,

I am trying Trinotate for my annotation. However, I found that the link https://data.broadinstitute.org/Trinity/__deprecated_trinotate_resources/Trinotate_v3_RESOURCES/Trinotate_v3.sqlite.gz is dead so we cannot obtain this SQLite boilerplate file needed for analysis.

Could you provide the Trinotate_v3.sqlite boilerplate or make it built-in somehow?

Gratefully,
Malik

Welcome @mmnashrullah

Yes, and the tool in Galaxy is also deprecated! So even if you are able to source this file, getting the tool to execute not will be possible, and we won’t be making any changes. The initial database was intended to be downloaded, not uploaded. Why? At the public Galaxy servers the sqlite artifacts involve special handling. So even if you found a suitable file, I wouldn’t expect you to be able to load it up, since that was never the original usage, plus some of the other dependencies are also deprecated!.

This report is from one of the last times I tried to get it to work and the job didn’t get far enough along to capture the template (my other copies are long gone) → Bug: Trinotate (Galaxy Version 3.2.1) · Issue #3875 · galaxyproject/tools-iuc · GitHub and this was the root problem → trinotate does not work, because GO FTP is gone · Issue #3555 · galaxyproject/tools-iuc · GitHub (never resolved, because the tool was deprecated by the authors).

If you wanted to set up a private Galaxy server, you could try, but the customization would be non-trivial (host the sqlite file somewhere, define the datatype). Also, making changes to the tool wrapper to do something custom (ignore missing dependencies, consume other changes with the datatype) might be possible but how difficult is hard to guess.

This is what else I could find:

Original development repository for what you see hosted at the public Galaxy servers (tool wrapper around the underlying tool) → tools-iuc/tools/trinotate at main · galaxyproject/tools-iuc · GitHub.

The source code for the underlying Trinotate tool (I would expect this to be complete, but you’ll need to check) → Releases · Trinotate/Trinotate · GitHub. Running this directly (in the correct containerized environment since dependencies will be important) might be your best option.

For updated tool options, maybe review the updated protocols involving Bioconductor tools?

The SQLite database was just a framework for a set of flat files joined by common keys and assigning annotation to transcriptomes is possible with other methods. Coming up with something custom in a workflow should be able to replicate what this older method was doing, and could function about the same way for a scientist: load data, launch, process, results.

I know this is probably disappointing but still hope this provides some clarity about the status and some ideas about alternatives. :slight_smile:

Thank you for your answer and clarification @jennaj

Currently I am setting up my own Trinotate and Trinity in my computer (under Windows 11 WSL) with the same version as installed in Galaxy Europe and finally run the perl script build_trinotate_boilerplate_sqlite_db.pl (which I found inside Trinotate source code and make some modification on the script). Now, I can get generate my SQLite boilerplate file and I will test it to the Galaxy Europe by uploading it.

Thank you for your attention.

1 Like