statonlab / galaxy-workflows Goto Github PK

Here I'm referring to the variant workflow linked in the README to the one in Ming's repository and not the one here. I'm assuming that one was meant to be here but didn't get pushed?

The very last step expect the user to enter the SnpEFF genome annotation database file. However, there are over 40K files and the only way to know the proper name of the database file to use is to use the command-line tool to list them. That's not possible in the workflow. So, there really isn't a way for the user to know what to put in that field.

Some potential fixes:

1. Create a web page on the Tripal site, that we can link to via the step instructions that list the genome names for species that are relevant to the site. (easiest)
2. Create a new input tool for Galaxy that provides an autocompete dropdown. In the tool settings we can cut-and-paste the list of databases, and let the user pick it with the tool via the workflow. I don't know how hard this would be. (hardest, but maybe helpful to lots of other workflows)
3. Download the database files from the SnpEff repository that are relevant for the species on the website and provide those as site-wide files. Then let the user specify the one to use in the workflow. We can then change the SnpEff tool in the workflow to use a file from the history. (requires reworking of the workflow)

The WGCNA workflow expect the user to enter the power to raise the data, but unfortunately, this can't be determined until they view the results from the previous steps. This could potentially be resolved by writing a small bit of R code to check on the output and make the determination for the user rather than have them enter it.

It seems that some of the workflow links in the README are pointing to raw workflow files in Ming's GitHub forked repository, and I noticed that the variant alignment provided here is not the same as the one linked in the README. The workflow on Hardwoods seems to be the one from Ming's GitHub.

I have two workflows that I have reworked a bit to provide more user-friendly messages. The most recent version of the Tripal Galaxy module can use the annotations in the tools in the workfows so folks don't have to edit them in the Tripal webform editing interface. But I can't submit the variant workflow here becuase the one i edited was the one from Ming's repository, which is linked from the README on this repository.

If we can sort out the confusion I can submit a PR with the updated workflows.