Use tumor ploidy and cellularity estimates from FACETS and/or Sequenza for clonality assessment

[susannasiebert]

Providing outputs from either of these tools (estimated cancer cell fraction - CCF) could be used to overwrite the current behavior for tumor clonality assessment in the aggregate report creation tiering.

More info from @mhoang22:
We know that if a SNP is in a region which is amplified (a region with copy number alteration), then its VAF might be high, but it’s not necessarily seen in so many cancer cells. This is the ploidy component.
We also know that biopsy is not pure, and might be a mix of tumor and normal cell. We have a purity number somewhat accounted for in pvac.
Once can use the combination of ploidy estimation, purity, and VAF to calculate the cancer cell fraction of each individual mutation: the fraction of cancer cells actually harbor the mutation. CCF close to 1 indicates that the mutation is likely to be in a founding clone.
The word ‘cellularity’ is used for either purity or cancer cell fraction, depending on the tool. Im not sure about sequenza, but for FACETS, the cellular fraction output is essentially CCF.

[malachig]

One way we could think about implementing this is to allow users to incorporate per variant cancer cell fraction (CCF) values instead of VAF. The CCF info could come from various tools but then get encoded into the VCF (e.g. as we do with VAtools already for RNA VAF for example).