search

galaxyproject / tools-iuc

Tool Shed repositories maintained by the Intergalactic Utilities Commission
https://galaxyproject.org/iuc
MIT License
164 stars 439 forks source link

Add Dimet tool suite: Differential Isotope-labeled targeted Metabolomics #5397

Open bdartigues opened 1 year ago

bdartigues commented 1 year ago

DIMet is a bioinformatics pipeline for differential analysis of targeted isotope-labelled data.

DIMet supports the analysis of full metabolite abundances and isotopologue contributions, and allows to perform it either in the differential comparison mode or as a time-series analysis. As input, the DIMet accepts three types of measures: a) isotopologues’ contributions, b) fractional contributions (also known as mean enrichment), c) full metabolites’ abundances. Specific functions process each of the three types of measures separately.

Note: DIMet is intended for downstream analysis of tracer metabolomics data that has been corrected for the presence of natural isotopologues. Make sure you that the metabolomics platform provides you the output of the correction procedure before using this pipeline.

hechth commented 1 year ago

Hi @bdartigues, it would be great if you could put the package first into conda. Please let me know if you would need any help with this! Just FYI, we are organizing regular Galaxy Metabolomics Community calls and you can sign up for the mailing list here. Do you already have some workflow in which you would like to use DIMet?

Best, Helge

hechth commented 1 year ago

@bdartigues I had a look at the package structure and will open up a PR shortly - maybe we can arrange a call to discuss the use of the package, how you use it in your lab to get an idea of how it would best be implemented inside Galaxy?

bdartigues commented 1 year ago

Dear @hechth, Thank you for your call proposal but I am not the author of the tool just the one in charge of integrating it into galaxy europe. I have already installed 3 tools on galaxy europe (Syndiva, Virhunter and decontaminator). This time, after asking Bjoern Gruening, we opted for a suite of tools because there are different stages and analyses in the tool. I've almost completed the tests on Galaxy local with planemo test, lint and serve. I thought I'd do the pull-request at the end of the afternoon. To be honest, we're going to be presenting the tool at the ECCB/IMSB on 27 July and it would be great if it could be available on Galaxy Europe by then. That's why I preferred to test the tool thoroughly before making my PR. Anyway, thanks in advance for all your help.

bdartigues commented 1 year ago

I saw that you added comments and PR to the DiMet repo, Many thanks for this but this version is not supposed to be changed, it has been tested deeply and this is the command line version. I will add my version this afternoon in tools-iuc. This way you can make comments on it.

hechth commented 1 year ago

Why is this version not supposed to be changed?

I had a look into the source code and more than half of the modules are not covered by unit tests. Maybe you tested them but it is not reproducible. It would be better if (1) all functionality would be covered by unit tests, (2) the package would be released with proper versioning and (3) it would be available via PyPI and bioconda to make it easy for users to install. This will make wrapping the galaxy tool easier and it will make the tool easier to maintain. Otherwise the whole code is not maintainable and will be soon deprecated.