Terms to describe transmission/inheritance of variance (from DECIPHER)

[mbrush]

Pasteing text from 1-19-18 email thread between Daniel Perrett, Melissa Haendel, and Peter Robinson:

---

Dear everybody
Daniel Perrett from the DECIPHER team is asking about terms to describe the transmission (inheritance) of variants. I think this might be best done with GENO.
I agree, Matt has been collecting similar requests from ClinGen and other clinical annotation groups.

Regarding where to put things, generally my feeling is that if it has to do with:
variation/inheritance/genotype and is applicable across species => GENO
human phenotype => HPO
sequence features in any species => SO

One can always import what you need into any given context from these and other ontologies.

Daniel, Matt is working on a manuscript for GENO and we would welcome suggestions! The tracker is here.

GENO currently has some inheritance terms that describe clinical inheritance (that in my opinion do no belong in GENO, they belong in HPO).
=> we can move them as needed, but if it is applicable in other species then probably needs to be in GENO so can be accessible for veterinary applications.

In contrast, GENO does not seem to currently have terms to describe transmission of variants -- this would be really useful for MME and other scenarios. Could we start a conversation about this?
Indeed! yes we need these, others asking for them too.

Cheers,
Melissa

The following text is by Daniel.
-Peter

In DECIPHER, the terms we accept for inheritance on a variant are:

• De novo constitutive
• De novo mosaic
• Paternally inherited, constitutive in father
• Paternally inherited, mosaic in father
• Maternally inherited, constitutive in mother
• Maternally inherited, mosaic in mother
• Biparental
• Imbalance arising from a balanced parental rearrangement
• Mitochondrial Homoplasmy
• Mitochondrial Heteroplasmy
• Unknown

We no longer collect the following terms, but they exist in the database

• CNV (= common in population)
• Inherited from normal parent
• Inherited from parent with similar phenotype to child
• Inherited from parent with unknown phenotype

These are a single field, so a patient only gets one of the above options.

Sequence ontology has the following terms:

• de_novo_variant
• germline_variant
• maternal_variant
• paternal_variant
• pedigree_specific_variant
• population_specific_variant
• somatic_variant

My current tentative thinking is that it would be useful to add the following terms:

• Germline variation (i.e. catch-all 'this is inherited' term as opposed to the de novo/somatic mutation term)
• Maternally inherited variation
• Paternally inherited variation
• Biparentally inherited variation
• Imbalance arising from a balanced parental rearrangement

And possibly also the following, matching the SO terms.

• Pedigree-specific variation
• Population-specific variation

But there are lots more we could add, e.g. Mitochondrial somatic variation

However, I do not know a lot about how others use HPO for inheritance. I thought it would be useful to ensure this is following a consistent and useful overall approach before requesting lots of terms in the issue tracker (happy to use that if you prefer, though).

I hope all is well with you

Daniel

[mbrush]

possibly relevant to #22 and #2