Isolate slots: conditional-requirement guidance, recommended/required cleanup, and isolate_meth replacement

[turbomam]

Several isolate slots were missing the "required for ..." guidance that the JGI form provides in parentheses for conditionally-required fields, which @aclum flagged in the #squad-isolates Slack thread. This adds that guidance to the comments of nuc_acid_absorb1, nuc_acid_absorb2, reference_genome, and replicate_group.

These conditions key on the JGI Sequencing Product, so the schema can't enforce them. They're guidance only. No range or constraint changes.

A follow-up will address the other part of her ask: descriptions/comments that are too verbose or inaccurate.

---

Update (scope grew during review). Beyond the comment guidance above, this PR now also makes enforcement changes agreed in #457 and in review:

• recommended dropped from the two fungal-draft slots (isolate_fungal_16s_screening, isolate_its_match_unite) and from reference_genome; kept on replicate_group and the ribosomal slots.
• collection_site_or_growth_conditions is now required on JgiIsolateTranscriptomeInterface (and no longer recommended on the genome interface).
• isolate_meth removed in favor of the existing MIxS dna_isolate_meth / rna_isolate_meth slots (required on their interfaces); the duplicated slot listings were de-duplicated.
• Test fixtures and example data updated accordingly, and lorem ipsum placeholders replaced with real values.

These change submission validation behavior, so the "no range or constraint changes" note above no longer holds.

[github-actions]

PR Preview Action v1.8.1
🚀 View preview at https://microbiomedata.github.io/submission-schema/pr-preview/pr-455/
Built to branch gh-pages at 2026-06-17 22:42 UTC. Preview will be ready when the GitHub Pages deployment is complete.

[Copilot]

Pull request overview

Updates the NMDC submission schema’s isolate-related slot metadata to include JGI’s conditional “required for …” guidance as slot comments, providing submitters with form-aligned instructions that the schema cannot enforce as constraints.

Changes:

• Added conditional-requirement guidance to replicate_group, nuc_acid_absorb1, nuc_acid_absorb2, and reference_genome slot comments in the base schema.
• Regenerated derived artifacts (expanded schema YAML, JSON project output, Python datamodel header timestamps) to reflect the new comments.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
src/nmdc_submission_schema/schema/nmdc_submission_schema.yaml	Regenerated schema YAML reflecting newly added conditional “required for …” slot comments.
src/nmdc_submission_schema/schema/nmdc_submission_schema_base.yaml	Source-of-truth change adding JGI conditional-requirement guidance to the relevant isolate slots’ comments.
src/nmdc_submission_schema/datamodel/nmdc_submission_schema.py	Regenerated datamodel (timestamp/header update) consistent with schema regeneration.
project/json/nmdc_submission_schema.json	Regenerated JSON output including the new slot comments arrays.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[aclum]

I am going over the reviewing the requirements for the first two rows.
isolate_ribosomal_seq is missing information about ITS requirements " ITS sequence must be >450 nt with <2% Ns"

[aclum]

I missed this before but First ribosomal sequence and First ribosomal sequence type are conditionally required, products that are not draft genomes (e.g. reseq) don't require this.

[turbomam]

now includes src/data/problem with examples of submissions where the pure LinkML validation doesn't cover JGI's conditional rules

[turbomam]

Per CONTRIBUTING.md ("This repository commits its generated files"), the regenerated build artifacts (the generated schema/nmdc_submission_schema.yaml, project/json, project/jsonschema, and datamodel/*.py) are committed alongside the source changes. Noting the choice explicitly since some recent PRs have included the generated files and others haven't.

[turbomam]

I am going over the reviewing the requirements for the first two rows. isolate_ribosomal_seq is missing information about ITS requirements " ITS sequence must be >450 nt with <2% Ns"

isolate_ribosomal_seq now carries the ITS threshold in both its description and its user-facing comment

ITS sequence is required for all samples for fungal drafts for strain verification. ITS sequence must be >450 nt with <2% Ns

Also keeps the other product thresholds (>1300 nt for microbial drafts, >220 bp with <10% Ns for single cells).

Nothing's checking the proportion of N bases or conditional lengths. I think we will have to do that in

• Extend TypeScript validation to support analysis_type-conditional required fields for isolate submissions nmdc-server#2177

[turbomam]

I missed this before but First ribosomal sequence and First ribosomal sequence type are conditionally required, products that are not draft genomes (e.g. reseq) don't require this.

Both isolate_ribosomal_seq and isolate_ribosomal_seq_type now note that JGI requires the ribosomal sequence for draft genome products but not for resequencing, methylation, or transcriptome products.

We can't express that as a LinkML rule because the requirement keys on the JGI Sequencing Product, which the schema doesn't carry as a slot, so it stays as guidance in the slot comments for now

see also

• Extend TypeScript validation to support analysis_type-conditional required fields for isolate submissions nmdc-server#2177.

[Copilot]

Pull request overview

Copilot reviewed 44 out of 44 changed files in this pull request and generated 5 comments.

Comments suppressed due to low confidence (2)

config/nmdc_schema_import.yaml:1222

• dna_isolate_meth is now injected into JgiIsolateGenomeInterface via import destinations, but it is also explicitly listed in JgiIsolateGenomeInterface.slots in the base schema. In the generated schema, this results in dna_isolate_meth appearing twice in the class slots list (and duplicate validation code in the generated datamodel). Ensure the slot is introduced exactly once (either via import destinations or via the base class slot list) and regenerate artifacts.

          slot_group: mixs_core_section

  - slot: dna_isolate_meth
    source: Biosample
    destinations:
      - JgiMgInterface
      - JgiMgLrInterface
      - JgiIsolateGenomeInterface
    modifications:
      - replace:
          rank: 16
          slot_group: jgi_metagenomics_section
          required: true
          recommended: false

config/nmdc_schema_import.yaml:3832

• rna_isolate_meth is now injected into JgiIsolateTranscriptomeInterface via import destinations, but it is also explicitly listed in JgiIsolateTranscriptomeInterface.slots in the base schema. In the generated schema, this results in rna_isolate_meth appearing twice in the class slots list (and duplicate validation code in the generated datamodel). Ensure the slot is introduced exactly once (either via import destinations or via the base class slot list) and regenerate artifacts.

  - slot: rna_isolate_meth
    source: Biosample
    destinations:
      - JgiMtInterface
      - JgiIsolateTranscriptomeInterface
    modifications:
      - replace:
          rank: 16
          slot_group: jgi_metatranscriptomics_section
          required: true
          recommended: false

[aclum]

@pkalita-lbl is there a way to make the docs preview work? The link above gets me a 404 (https://microbiomedata.github.io/submission-schema/pr-preview/pr-455/)

[pkalita-lbl]

is there a way to make the docs preview work?

@aclum Not easily. The workflow that builds the docs site to include the DataHarmonizer playground isn't readily compatible with the way the PR preview actions expects the workflow to be setup. We might be able to make it work, but I don't know what that would look like off the top of my head.

[aclum]

okay, no problem.