Implement variant sets to support genotypes and other combinations of variants

[malachig]

Currently we have several issues that could be resolved by supporting a notion of variant sets.

1. A combination of variants that co-occur in a single individual or tumor. This could be called a Genotype variant set.
2. A combination of variants that are deemed equivalent to each other. This happens everywhere that we have an amino acid change where multiple nucleotide changes can cause the same amino acid change. Right now we pick a representative. Variant sets would allow us to enumerate them and get the connection to ClinGen Allele Registry, COSMIC, etc. It would also help other connect their variants to CIViC evidence and assertions. This could be called Alternates variant set. Or Equivalents variant set.
3. Our current idea of variant groups could be rolled into this. This could be called Conceptual variant set.

Would we try to support ploidy status of alleles?