Add optional is_aberration and is_balancer flags to Allele

[sjm41]

Summary

FlyBase makes a clear distinction between traditional single-gene alleles and chromosomal aberrations (multi-gene alleles) — handling them as separate data classes (FBal vs FBab) — and balancers are a subtype of aberration. This adds two optional booleans to Allele/AlleleDTO to capture that distinction within the Alliance single Allele class:

• is_aberration — indicates the allele is a chromosomal aberration (multi-gene allele) rather than a traditional single-gene allele.
• is_balancer — indicates the allele is a balancer (a recombination-suppressing chromosomal rearrangement). Balancers are a subtype of aberration, so this is only applicable to alleles for which is_aberration is true (documented in the slot description; not schema-enforced).

Both follow the existing is_extinct / is_extrachromosomal / is_integrated pattern — plain optional booleans reused directly in the DTO.

Also

Removes the dead aberration slot_usage entry (flagged by [CG] as no longer declared) and its unused standalone slot definition, which is_aberration supersedes. Both were already absent from the generated artifacts.

Notes

• No LinkML rule enforcing balancer ⟹ aberration (considered, deliberately left out).
• generated/ artifacts will be regenerated by CI.

🤖 Generated with Claude Code

[claude]

Review

Adds two optional boolean flags is_aberration and is_balancer to Allele, and removes the stale, unused aberration string slot (which carried a 2022 "not being used" TODO).

The change is clean and follows the established is_extinct/is_extrachromosomal/is_integrated pattern exactly:

• Slots added to both Allele and AlleleDTO slot lists.
• For plain booleans the DTO correctly reuses the same slot name (no _curie/_name suffix needed).
• domain: Allele, range: boolean, clear descriptions, and required: false in slot_usage — matching the sibling flags.
• The removed aberration slot has no remaining references in model/schema/** or test/data/**.

No assertions to raise.

Minor note (non-blocking): the is_balancer description states balancers are a subtype of aberration and only applicable when is_aberration is true. That dependency is documentation-only and not enforced by the schema — fine to leave, just flagging that nothing prevents is_balancer: true with is_aberration unset/false.

LGTM.

[sjm41]

FYI: I put this proposal on the agenda for the Disease and Pheno WG meeting on July 8th.

[gm119]

I would change the 'is_aberration' descriptions so that it avoids mention of 'a multi-gene allele' as a single transgenic insertion can also affect multiple genes (e.g. insertion into overlapping genes), and we would not regard those as an 'aberration'.

Our current FB QA/QC checks require us to type aberrations with a SO term that sits in the 'chromosome_structure_variation ; SO:1000183' branch, this term is defined as 'An alteration of the genome that leads to a change in the structure or number of one or more chromosomes.'

So I think the key part is the 'change in the structure or number of one or more chromosomes'

So I would use the SO term as the basis of the linkml definition for is_aberration instead of having 'a multi-gene allele'., so maybe:

is_aberration:
description: >-
Indicates whether the allele is a chromosomal aberration, defined as 'An alteration of the genome that leads to a change in the structure or number of one or more chromosomes.' (SO:1000183).
domain: Allele
range: boolean

(or however SO terms should be referred to)

[gm119]

looks good