Use cases/requirements gathering

[jsstevenson]

Initial stabs at some user stories to inform additional requirements and prioritization:

• Easy-to-stand-up variant registration service for a research group. Emphasis would be placed here on documentation, quickstart options (eg a Makefile, dockerization, etc), access to more advanced normalization functions Variation Normalizer integration #19 . Potentially some craftier search options (which might strain our adherence to the MutableMapping datastore interface). Maybe an option to provide ingest (and even annotate?) VCFs. This is probably closest to who our group plans to be developing for over the next few months.
• Caching tool to increase throughput for research/curation pipelines that want to VRSify their variation objects. This probably doesn't require much in the way of lookup options beyond dereferencing VRS digests, but would want fast read access (hence Redis support).
• Public-facing variant registration service. Possibly entails some differentiation of users (write endpoints require authorization, public access is read-only)

I think this project is intended to be pretty light and modular, so some of these needs may be out-of-scope.