This still needs to be filled out: how will the ECC be assigned, what will the permissible values be, and what do each of the permissible values mean.

Thank you for your comments. However, I am not sure what the description of the binary feature refers to.
In my understanding, it could be one of two cases:

1. It describes the ECC, such as how we evaluate or confirm that ECC - criteria to define this ECC, for example:
   descriptions:
   "true":
   summary: Confirm by Sanger sequencing.
   "true":
   summary: confirm by RNA seq.

2. It describe the ontology characteristics, meaning it should be in the definition of an ontology, for example:

Ontology Class: Diffuse midline glioma, H3 K27-altered
ECC:
- identifier: ECC-MOLEC-00001
name: H3 K27M Mutation
appear: true
- identifier: ECC-MOPHL-00001
name: Vascular Proliferation
appear: false
...

Please let us know your ideas or examples if possible, and correct me if I misunderstood.

I submit we should center on the name ("" alteration) for point mutations.

I'm not going to suggest changes for every instances of mutation -> alteration, but please make those (and use alteration in the future).

Thanks for the suggestions, I'd like to note that alteration term is a board term in genetics definition, which can be refer to various type of DNA changes including stucture variations, deletions, insertions, and more; not just a point mutation. Personally, I think that mutation" or "mutant would be more percise in the context. These terms are commonly used in scientific papers. That said, I also understood that we might want to generalize our terminolgy, so let's discuss futher to find the best convention for us to avoid ambiguity.

I lean towards mutation when the variant is limited to SNV and/or Indel such as in this case. When the type of hit can be more broad, then alteration.

How about "The H3 K27M mutation is present in the sample."

I lean towards mutation when the variant is limited to SNV and/or Indel such as in this case. When the type of hit can be more broad, then alteration.

This is a very important comment. I was going to suggest it if it wasn't in here.

How about "The H3 K27M mutation is present in the sample."

I would think that for most modern datasets, we would be basing the call purely on NGS. The way this is written, it sounds like NGS would only be used for confirmation.

Overall, I don't really like the details section here. The way I think of it is that we want the semantics to be clear, i.e. it's important that people know what true means and what false means. I think that's abundantly clear from the summary alone. However, different institutions may have different mechanisms and even possibly different standards for deciding the presence/absence of a variant, and I think it's OK (and even preferable) to allow that variation. And it can change over time. The more detail you give about what we really mean about presence and absence, the more in sync everybody will be in theory, but it's also more work to maintain the characteristics, and I think by being so proscriptive we could find that it opens up a lot of debate that we would be better to avoid by just leaving those details up to each user. In this version we partly specify the criteria for determining presence/absence, but we leave a lot of it open to interpretation (e.g. variant calling methods and criteria on the NGS data). I think it's probably the worst of both worlds.

My suggestion is that we omit the information currently in details entirely, and I would say we shouldn't need value details for Booleans generally (and possibly not for some others as well). If the guidance that we put in here is from a referenced resource, then you could move this information to the "context" of that reference, where you could possibly summarize it as "This paper recommends practices for determining the presence of H3 K27M using immunohistochemistry and NGS testing".