Arvados

What I need to preserve are the metadata of the input. Let's get a bit more concrete to explain better:

We have a workflow that processes RNAseq data. It creates an output collection with many folders / files.

Now what we have is a second workflow to process data for a downstream application. This workflow pulls out certain filetypes from the output collection of the first, like .gct files. Now what we want to add is a file that captures the metadata associated to these .gct files, in the form of the workflow uuid of the first processing workflow. So the idea was to take that from the first output collection (which is input for the second workflow) in the forms of metadata. The problem is if the workflow was run multiple times, I would need the uuid of the output collection to identify the one with the correct metadata.