search

biocore / metagenomics_pooling_notebook

Jupyter notebooks to assist with sample processing
MIT License
8 stars 16 forks source link

SampleSheet generator module runs with warnings. Well_description column broken. #46

Closed RodolfoSalido closed 2 years ago

RodolfoSalido commented 2 years ago

After cloning the current repository and running in JupyterHub, I ran into some runtime warnings around the SampleSheet generator module.

Here are the watermarks:

Screen Shot 2021-10-27 at 11 35 54 AM

The first warning doesn't recognize the 'Experiment' key in the metadata dictionary: Screen Shot 2021-10-27 at 11 33 01 AM

After deleting the 'Experiment' key and typing in a 'experiment_design_description', I still get a warning about the 'Well_description' column being empty:

Screen Shot 2021-10-27 at 11 34 16 AM

Indeed, after opening the samplesheet file we find that 'Well_description' is empty:

Screen Shot 2021-10-27 at 11 40 31 AM

ElDeveloper commented 2 years ago

Thanks for reporting this @RodolfoSalido. My understanding is that Well_description is supposed to be populated with the intended Qiita sample nameso. @RodolfoSalido @mmbryant23 @antgonza Do you think it would be good enough to automatically scrub the sample ids (to meet Qiita's requirements - only alpha numeric characters and periods) and put those values there? Generally speaking this will be the key that we'll want to use to link to the metadata sheets that our collaborators will send us.

Additional background: previously the Well_description column was populated by someone (not by the notebooks themselves) and so once the sample sheets made their way to barnacle we could use that information to fill the sample name in the Qiita prep files. Since we don't really want manual modifications to the sample sheet, I think the best would be to repopulate this column with some useful value.

Note that Qiita and Illumina have incompatible character restrictions, so we can't reuse the Illumina sample name here. Maybe the way to go is to take the original (unscrubbed) sample name and scrub that for Qiita only?

@RodolfoSalido As for the Experiment key, that looks like it was my bad when merging some of @ahdilmore's changes. I've fixed that here #47. Feel free to review/merge 👍

antgonza commented 2 years ago

IMOO scrubbing for Qiita and making sure there are no duplicated values should be a good enough solution.

ElDeveloper commented 2 years ago

Thanks for the suggestion @antgonza and for reporting this @RodolfoSalido! I just submitted #49