Document JSON.gz files, and Transcript format

[davmlaw]

https://github.com/SACGF/cdot/wiki/Transcript-JSON-format

Maybe break into the JSON.gz (ie all the keys) and the transcript details (which is in REST)

Be sure to document strangeness, ie:

In the JSON - we store the gene info using the ID (eg RefSeq 80167) as key

When we build the data from UTA (your example has "url": "postgresql://uta.biocommons.org/uta_20210129" - we can't know they ID, only the symbol. So we make up a fake symbol