Extend schema to cover metabolomics output to enable search by compound

[cmungall]

C-MS based metabolomics workflow output examples from @corilo here:

https://drive.google.com/drive/u/1/folders/1_dHFvIK9PwJCKVJznwqWgqvfOIhTCkvx

Expected changes:

Metadata JSON file: structure will change after the adoption of the labels defined on AIM 1,
DataTable CSV file: metabolite names will change (removing comments and extra fields), and more columns will be added to include CAS and KEGG compound ID

NOTE: for now, make comments on @dehays google doc https://docs.google.com/document/d/15fga30d619WRxAUk8LyrojwIN1m89_K-sIrmUg4Y3tY/edit rather that commenting in ticket. We will still use this ticket to track status

As a first pass we will not try and capture everything from a metabolomics workflow, just the aspects that are necessary for search

[dehays]

Yuri @corilo - will produce JSON document for a single execution of his workflow. Then get feedback (Chris, Bill, David) and possibly iterate before continuing to do remaining metabolomics workflow instances - structure will then be used for lipidomics and OM workflow execution metadata

[corilo]

@dehays, @cmungall @wdduncan
Here is the first draft of the JSON document. Please provide feedback on the structure, labels, and any missing data as required.

I forgot what the "was_informed_by" and what it should contain. Please help?

https://drive.google.com/file/d/1BnP8q-iDQP2vmswN9v68WDQfNG4uDek5/view?usp=sharing

Thanks

[dehays]

"has_input": [
            "emsl:sha256:37417faf2c1b07ef9c59868683e41577bb3a745128bdc88b6cc59e579b5b30d0"
        ]

This ID (the sha256 hash) doesn't match the ID of any instrument output EMSL has provided to NMDC. This uses that dataset ID (that some parts of EMSL use and some parts do not) with a prefix of "output_". Looks like "emsl:output_500097". We can revisit how EMSL sets unique IDs, but the IDs need to be consistent so that there is a path back from analysis to sample and study.

"was_informed_by" is a relationship on an analysis execution activity (i.e. instance of running the metabolomics analysis workflow) that refers to the instrument run (OmicsProcessing) entity. Again, the ID currently looks like "emsl:500097" and uses the dataset ID of the instrument run.

The Metabolites object - I was expecting an array rather than an object, but a more important conversation is in how this will relate to the structure @cmungall is describing in #176 I hope to understand this better after speaking with Chris on Wednesday