Add HRA spleen NTR batch outputs and review reports

[zhengj2007]

Generated via bulk_ntr_workflow from source_data/spleen.csv.

Pipeline completed through Stage 4:

• 9 new UBERON leaf terms in final template
• 0 group terms
• 1 confirmed existing match excluded
• 0 possible matches identified
• 0 out-of-scope terms
• 0 manual-curation items

QC status:

• 0 [PENDING] definitions
• 0 INFER/UNRESOLVABLE/NEEDS_MAPPING markers

Includes:

• Template and groups TSVs in src/templates/
• Detailed review reports: candidates, errors, input, name_corrections
• Source data snapshot in bulk_ntr_workflow/source_data/

Signed-off-by: @dragon-ai-agent

[ai4c-agent]

Claude finished @zhengj2007's task in 5m 5s —— View job

---

PR Review: Add HRA spleen NTR batch outputs and review reports

Checklist:

• Hierarchy and parents are consistent
• Definitions and supporting references look adequate
• Obsolete or replacement handling is safe (N/A)
• OBO and ODK conventions appear preserved (with caveats — see below)
• No obvious CI, auth, or workflow regression

---

Summary

This PR adds ROBOT template outputs for 9 new spleen anatomy terms (IDs UBERON:8930191–UBERON:8930199), allocated to Jie Zheng in idranges.owl (idrange:42). Pipeline stages 1–4 completed cleanly: 0 errors, 0 pending definitions, 1 confirmed existing match excluded (Connective tissue → UBERON:0002384). References are generally well-supported with multiple DOIs/PMIDs.

Several issues need attention before merge or before the template is applied via ROBOT:

---

🟡 IMPORTANT Issues

1. Malformed DOI URL in arterial capillary def_xref

src/templates/hra-spleen.template.tsv, line 3 — the first DOI contains spaces:

https://doi.org/ 10.1182/ bloodadvances.2017015073

This URL is invalid and will produce a malformed xref in the ontology. Correct form: https://doi.org/10.1182/bloodadvances.2017015073 (also duplicated correctly later in the same cell as DOI:10.1182/bloodadvances.2017015073). The malformed version originates from the source data (spleen.csv line 3) and carried through unchanged.
Fix this →

2. Space in PMID reference for Superficial Mantle Zone

src/templates/hra-spleen.template.tsv, line 10 — def_xref contains PMID: 20126587 (space after colon). This should be PMID:20126587 with no space per standard OBO xref formatting.

3. Non-standard capitalisation in LABEL values

UBERON convention uses lowercase for anatomical term names (compare periarterial lymphatic sheath, splenic sinusoid, white pulp of spleen). The following labels violate this:

Current label	Suggested label
Inner Mantle Zone	inner mantle zone
Superficial Mantle Zone	superficial mantle zone
Ring Fiber	ring fiber
Splenic Red Pulp Stroma	splenic red pulp stroma
Inner PALS	inner periarterial lymphatic sheath
Outer PALS	outer periarterial lymphatic sheath

The pipeline's name_corrections.tsv caught three labels (Arterial Capillary, White Pulp Lymphatic Vessels, Subcapsule) but missed these five. The original capitalised forms should be added as synonyms.

4. "Inner PALS" / "Outer PALS" — abbreviation should be expanded

"PALS" stands for periarterial lymphatic sheath (= UBERON:0001960). Using an abbreviation as the primary label is not UBERON convention. The labels should be inner periarterial lymphatic sheath and outer periarterial lymphatic sheath, with PALS retained as a synonym.

5. Ring Fiber is_a UBERON:0006804 (reticular tissue) — potential classification error

UBERON:0006804 is defined as a type of tissue (reticular connective tissue). A ring fiber is an individual structural fiber, not a tissue type. Classifying ring fiber is_a reticular tissue is logically imprecise — a ring fiber is a component of reticular tissue, not a subtype of it. UBERON does not appear to have a dedicated "reticular fiber" class, so this may require either:

• Using UBERON:0000061 (anatomical structure) as a fallback is_a, with part_of UBERON:0003910 (splenic sinusoid) preserved, or
• Curator review to identify a more appropriate structural parent.

---

🔵 Suggestions

6. arterial capillary — no part_of relationship

The term is defined in general terms (not spleen-specific), with only is_a capillary. If this is intentionally a general capillary subtype term, that is fine, but the definition's example context of the spleen is not captured in any relationship. Consider noting in the definition or adding a Present_in_taxon annotation that this applies to mammalian vasculature generally. (If the term is meant to be spleen-specific it should have a part_of red pulp of spleen relationship.)

7. Stage 5 (ODK registration) is pending

The templates are staged in src/templates/ but have not yet been registered in src/ontology/uberon-odk.yaml. This must happen before the ROBOT build can use them. Consider whether this should be part of this PR or a follow-up.

8. ASCTB-TEMP IRIs in def_xref

All terms include https://purl.org/ccf/ASCTB-TEMP_* URIs in the def_xref column. Per the workflow's own QC checklist, these do not count as bibliographic references. They are acceptable as cross-references but should be moved to the xref column rather than def_xref to keep the definition citation column clean.

---

Parent term verification

All UBERON parent IDs resolve correctly:

Term	is_a	part_of
arterial capillary	UBERON:0001982 capillary ✓	—
Inner/Superficial Mantle Zone	UBERON:0001744 lymphoid tissue ✓	UBERON:0010421 spleen B cell corona ✓
Inner/Outer PALS	UBERON:0000064 organ part ✓	UBERON:0001960 periarterial lymphatic sheath ✓
Ring Fiber	UBERON:0006804 reticular tissue ⚠️	UBERON:0003910 splenic sinusoid ✓
Splenic Red Pulp Stroma	UBERON:0003891 stroma ✓	UBERON:0001250 red pulp of spleen ✓
subcapsular zone	UBERON:0000064 organ part ✓	UBERON:0002106 spleen ✓
white pulp lymphatic vessel	UBERON:0001473 lymphatic vessel ✓	UBERON:0001959 white pulp of spleen ✓

Recommendation: REQUEST CHANGES — The malformed DOI URL (issue 1), capitalization issues (issues 3–4), and the potential is_a misclassification for Ring Fiber (issue 5) should be resolved before this template is used to generate ontology terms.

[ai4c-agent]

Several issues need attention before this template is applied:

🟡 IMPORTANT:

1. Malformed DOI URL in 'arterial capillary' def_xref: 'https://doi.org/ 10.1182/ bloodadvances.2017015073' has spaces and is invalid.
2. 'PMID: 20126587' has a space after the colon in Superficial Mantle Zone def_xref.
3. Multiple LABEL values use non-UBERON capitalization: 'Inner Mantle Zone', 'Superficial Mantle Zone', 'Ring Fiber', 'Splenic Red Pulp Stroma', 'Inner PALS', 'Outer PALS' — all should be lowercase.
4. 'Inner PALS'/'Outer PALS' use an abbreviation; labels should be expanded to 'inner periarterial lymphatic sheath' / 'outer periarterial lymphatic sheath'.
5. Ring Fiber is_a UBERON:0006804 (reticular tissue): a ring fiber is an individual fiber, not a tissue type — this is_a parent may be a logical misclassification.

See the full review in the PR comment for details and suggestions.

[github-actions]

This PR has not seen any activity in the past month; if nobody comments or reviews it in the next week, the PR creator will be allowed to proceed with merging without explicit approval, should they wish to do so.

[cellsemantic]

Reports relocated. Route: manual hand-edit (~11 terms, <20). TODO: convert to OBO stanzas; verify defs/PMIDs/parents.

---

Automated update (@cellsemantic / dragon-ai-agent, for @dosumis). Reports were relocated out of src/templates/ → bulk_ntr_workflow/outputs/<name>-reports/ to match the canonical layout (PR #3700). Note: the corrected reports-location docs currently live only on add-hra-muscular-ntr (#3700), not master. Batch-size policy for this HRA set: ≥20 new terms → ODK-register; <20 → hand-edit into uberon-edit.obo.

[ai4c-agent]

Claude finished @dosumis's task in 4m 9s —— View job

---

PR Review: Add HRA spleen NTR batch outputs and review reports

Checklist:

• Hierarchy and parents are consistent (parents resolve, with one classification caveat — see Ring Fiber below)
• Definitions and supporting references look adequate (with two formatting defects — see below)
• Obsolete or replacement handling is safe (N/A — no obsoletions in this PR)
• OBO and ODK conventions appear preserved (with caveats — see below)
• No obvious CI, auth, or workflow regression

---

Summary

Re-reviewed after the latest commit (d61ae4a), which only relocates the QC report artifacts from src/templates/hra-spleen-reports/ to bulk_ntr_workflow/outputs/hra-spleen-reports/ — a pure file move, no content changes. All substantive issues flagged in the original review (2026-05-26) remain unresolved in src/templates/hra-spleen.template.tsv — inline comments posted on each:

• 🟡 Malformed DOI URL with embedded spaces in the arterial capillary def_xref (line 3)
• 🟡 Space after colon in a PMID: def_xref for Superficial Mantle Zone (line 10)
• 🟡 Six LABEL values still use non-UBERON capitalisation (Inner Mantle Zone, Superficial Mantle Zone, Ring Fiber, Splenic Red Pulp Stroma, Inner/Outer PALS)
• 🔵 "Inner PALS"/"Outer PALS" should be expanded to full form with PALS kept as a synonym
• 🟡 Ring Fiber is_a UBERON:0006804 (reticular tissue): confirmed via uberon-edit.obo that UBERON:0006804 is a tissue class (a loose connective tissue subtype), while a ring fiber is an individual structural fiber — likely classification error

Verified there are 0 errors, 0 pending definitions, and all other referenced UBERON parent/part_of IDs resolve correctly (checked via obo-grep.pl).

Process note: a later thread comment (@cellsemantic, 2026-07-09, for @dosumis) establishes a batch-size policy: batches with <20 new terms should be hand-edited directly into uberon-edit.obo (per the terms/ checkout workflow in CLAUDE.md) rather than registered as an ODK ROBOT template component. This batch has 9 new leaf terms (well under 20), and that comment includes a TODO to convert this template to OBO stanzas. The PR currently only contains the ROBOT template in src/templates/ (Stage 5/ODK registration hasn't run, consistent with the policy) — but per the stated policy the end state should be hand-edited term stanzas in uberon-edit.obo, not a template destined for ODK registration. Worth confirming whether that conversion happens in this PR or as tracked follow-up before merge.

Recommendation: REQUEST CHANGES (review submitted) — the malformed DOI/PMID xrefs, capitalisation issues, and the Ring Fiber is_a classification should be resolved, and the manual-hand-edit conversion per the new batch-size policy should be confirmed, before this is merged.

[ai4c-agent]

🟡 IMPORTANT: The def_xref cell still contains a malformed DOI URL with embedded spaces: https://doi.org/ 10.1182/ bloodadvances.2017015073. This is invalid and will produce a broken xref when the template is applied. The correctly-formed DOI:10.1182/bloodadvances.2017015073 is already present later in the same cell, so this looks like a leftover duplicate from the source CSV rather than a needed reference — recommend removing the malformed variant. Also present at bulk_ntr_workflow/source_data/spleen.csv line 3.

Fix this →

[ai4c-agent]

🟡 IMPORTANT: PMID: 20126587 (space after colon) in the def_xref cell — standard OBO xref formatting requires PMID:20126587 with no space. The correctly-formatted PMID:20126587 is already present later in the same cell, so the malformed copy can simply be dropped.

[ai4c-agent]

🟡 IMPORTANT: Several LABEL values in this range still use non-standard capitalisation (UBERON convention is lowercase, cf. subcapsular zone on line 9 and white pulp lymphatic vessel on line 11 which are correctly lowercase):

Line	Current label	Suggested label
4	Inner Mantle Zone	inner mantle zone
5	Inner PALS	inner periarterial lymphatic sheath
6	Outer PALS	outer periarterial lymphatic sheath
7	Ring Fiber	ring fiber
8	Splenic Red Pulp Stroma	splenic red pulp stroma
10	Superficial Mantle Zone	superficial mantle zone

name_corrections.tsv caught three other labels in this batch (Arterial Capillary, White Pulp Lymphatic Vessels, Subcapsule) but missed these five — worth checking why the pipeline's capitalization-correction rule didn't fire here. The original capitalised forms should be retained as synonyms.

[ai4c-agent]

🔵 SUGGESTION: "PALS" is an abbreviation for periarterial lymphatic sheath (UBERON:0001960, which is already the part_of parent on both these rows). UBERON convention avoids abbreviations as primary labels — recommend inner periarterial lymphatic sheath / outer periarterial lymphatic sheath as the LABEL, with PALS kept as a synonym (folds into the capitalisation fix above).

[ai4c-agent]

🟡 IMPORTANT: is_a UBERON:0006804 (reticular tissue) for "Ring Fiber" is a likely classification error. I checked UBERON:0006804 directly in uberon-edit.obo:

[Term]
id: UBERON:0006804
name: reticular tissue
def: "Loose connective tissue which consists of an irregular network of cells and reticular fibers..."
is_a: UBERON:0011825 ! loose connective tissue

reticular tissue is a tissue class (a type of loose connective tissue). A ring fiber, per this row's own definition ("a circumferential reticular fiber that encircles a splenic sinusoid"), is an individual structural fiber — a component of reticular tissue, not a subtype of the tissue itself. ring fiber is_a reticular tissue is logically imprecise.

UBERON does not appear to have a dedicated "reticular fiber" class. Options:

• Fall back to a generic structural parent (e.g. UBERON:0000061 anatomical structure) with part_of UBERON:0003910 (splenic sinusoid) preserved, or
• Curator review to identify/mint a more appropriate parent (e.g. a "reticular fiber" class if warranted).

[ai4c-agent]

PR Review: Add HRA spleen NTR batch outputs and review reports

Checklist:

• Hierarchy and parents are consistent (parents resolve, with one classification caveat — see Ring Fiber below)
• Definitions and supporting references look adequate (with two formatting defects — see below)
• Obsolete or replacement handling is safe (N/A — no obsoletions in this PR)
• OBO and ODK conventions appear preserved (with caveats — see below)
• No obvious CI, auth, or workflow regression

Summary

Re-reviewed after the latest commit (d61ae4a), which only relocates the QC report artifacts from src/templates/hra-spleen-reports/ to bulk_ntr_workflow/outputs/hra-spleen-reports/ — a pure file move, no content changes. All substantive issues flagged in the original review (2026-05-26) remain unresolved in src/templates/hra-spleen.template.tsv:

• 🟡 Malformed DOI URL with embedded spaces in the arterial capillary def_xref (line 3)
• 🟡 Space after colon in a PMID: def_xref for Superficial Mantle Zone (line 10)
• 🟡 Six LABEL values still use non-UBERON capitalisation (Inner Mantle Zone, Superficial Mantle Zone, Ring Fiber, Splenic Red Pulp Stroma, Inner/Outer PALS) — inline comments posted with specifics
• 🔵 "Inner PALS"/"Outer PALS" should be expanded to full form with PALS as a synonym
• 🟡 Ring Fiber is_a UBERON:0006804 (reticular tissue): confirmed via uberon-edit.obo that UBERON:0006804 is a tissue class (loose connective tissue subtype), while a ring fiber is an individual structural fiber — likely classification error

Verified there are 0 errors, 0 pending definitions, and all other referenced UBERON parent/part_of IDs resolve correctly (checked via obo-grep.pl).

Process note: a later thread comment (@cellsemantic, 2026-07-09, for @dosumis) establishes a batch-size policy: batches with <20 new terms should be hand-edited directly into uberon-edit.obo (per the terms/ checkout workflow in CLAUDE.md) rather than registered as an ODK ROBOT template component. This batch has 9 new leaf terms (well under 20), and the same comment includes a TODO to convert this template to OBO stanzas. The PR as it stands still only contains the ROBOT template in src/templates/ (not yet ODK-registered, which is consistent with the policy — Stage 5 hasn't run) — but per the stated policy the end state for this PR should be hand-edited term stanzas in uberon-edit.obo, not a template destined for ODK registration. Worth confirming whether this PR will be converted, or whether that's tracked as follow-up work, before merging in its current form.

Recommendation: REQUEST CHANGES — the malformed DOI/PMID xrefs, capitalisation issues, and the Ring Fiber is_a classification should be resolved (and the manual-hand-edit conversion per the new batch-size policy should be confirmed) before this is merged.