Existing Wording: Genetic Interpretations can be "derived from" Genetic Assertions.

Genetic Impacts MUST be "derived from" Genetic Assertions.

Comment:

I believe that any phenotype analysis of genomics observation should be explicitly associated with the observations based on which the analysis was done, whether it?s called interpretation, impact, etc. The pragmatic level of requirement is SHOULD even though I agree with MUST but think that in reality it will be hard for implementers to actually realize this kind of associations, and thus this profile might set up a too high bar of compliance, at least for initial effrots (which seems to be always the case in genomics... .

Note that the general approach implied from those bullet points is similar in principle to the design principle of CG v3 and CDA specs, through the Clinical Genomics Statement model.

Summary:

can and must or should?