FR: Order of samples in MultiQC same as in samples

[Maarten-vd-Sande]

Is your feature request related to a problem? Please describe. The current semi-random ordering is ugly, and might hide obvious patterns.

[Maarten-vd-Sande]

This does not seem to be supported by MultiQC. The general table is sorted alpabetically. Modules is currently mildly unclear how the work, possibly alphabetically, possible in order files are fed to MultiQC. So we might be able to "hack" this by feeding our files in the order we want to show them in a module

[Maarten-vd-Sande]

Nope its alphabetically :sob:

[Maarten-vd-Sande]

Putting a wontfix on this. If something changes in the future in the way MultiQC handles this, would be really nice to add!

[siebrenf]

Two options:

• while fixing #471 I learned that set() sorts the data alphabetically. Should this be of influence here, you can use seq2science.utils.unique() instead.
• you could perhaps prefix the sample names with numbers (01_, 02_, etc...)

[Maarten-vd-Sande]

It's a MultiQC internal thing, so the first option does not do anything.

Yes prefixing works, but that means we would have to change a lot how we do stuff in the pipeline. Probably easier to change MultiQC source then (but at first glance, it's not that easy).