Opencrevat parsing

[wilkb777]

This is a complete overhaul of the project structure, parsing strategy, and which annotation source will be used for model generation and DITTO scoring closing #3 .

This work specifically implements a configurable (by file) parser for de-normalizing variant annotations from OpenCRAVAT into a single variant + transcript combination per line. This changes allows for efficient model generation and facilitates scoring variant + transcript combination individually instead of wrapping up the score just to the variant.

Testing was completed on an example data set
testing-vars.csv.zip and verified as parsing correctly using the test config included in the repo.