Consider an allelic requirement qualifier on Variant Pathogenicity Interpretations

[mbrush]

Should we consider an allelic requirement qualifier on variant pathogenicity interpretations that would allow us to capture how many copies of a variant would be required to cause the indicated disease??

Recall that the subject here can be single locus complements/genotypes (e.g. variant x/variant x, or variant X / WT) - which provides context from which we might infer an allelic requirement. But most VPI annotations are made on simple variants, and in cases where we have knowledge about whether one or two copies of the variant in question would be required to cause disease, we could allow this to be captured using an 'alelelicRequirementQualifier' attribute on the statement.

Values would be things like 'monoallelic', 'biallelic', etc . . . see slide 6 of the deck here for more

[larrybabb]

not as a requirement. If you use ClinVar and the ACMG guidelines as the gold standard to specify what we are modeling then I would not. This question has come up many times in the past. In the end, the experts seem to land on the side of qualifying the sequence change only, not it's allelic state as part of the qualification of the pathogenicity assertion. If the assertion associated to a recessive disease then the patient level application of the assertion may manifest to a Positive or Carrier patient level overall interpretation depending on whether the specific finding is heterozygous or homozygous. In the end, the variant (regardless of zygosity) is causal for the disease. There's a distinct separation of that assertion from the specific manifestation found in a given subject/patient. Let me know if you'd like to have Steven Harrison or Heidi Rehm discuss this in greater detail.

[larrybabb]

after looking at slide 6 of the Sanger deck you provided above. I would suggest that these allelic requirements are the requirements used when assessing the patient's findings in conjunction with the Var Path interp assertion. This is part of the conflation of "what is a variant?". When assessing a simple allele, we are merely assessing the impact that its change causes to the coding of the protein in isolation from whether or not that change exists in various allelic states. This "kind" of variant is a building block that ties us to an actual variant found in a genome with its full genomic "allelic state" (zygosity, copies, etc..). A representation of the simple allele in the context of a sample from a patient (GeneX c.500G>A heterozygous), is not the same as the representation of the simple allele in a reference copy of a chromosome, gene, transcript, rna, protein, etc.. in isolation (GeneX c.500G>A). These two representations of variants need to be clearly separated in order to clarify when we should be using one version (the reference allele/variant) versus the other (the patient/subject context of that variant). Even though the later must be directly linked and dependent on the reference representation.