fix: pre-existing metabolite identity, structure and cross-reference errors

.github/workflows/gene-essentiality.yml

@@ -425,7 +425,6 @@ jobs:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           PR_NUMBER: ${{ inputs.pr }}
           BASE_REF: ${{ inputs.base }}
-          CSV_URL: ${{ github.server_url }}/${{ github.repository }}/blob/${{ github.head_ref || github.ref_name }}/data/testResults/gene-essential-diff.csv
           SUMMARY_URL: ${{ github.server_url }}/${{ github.repository }}/blob/${{ github.head_ref || github.ref_name }}/data/testResults/gene-essential_summary.md
         run: |
           if [ -z "$BASE_REF" ]; then
@@ -444,7 +443,6 @@ jobs:
             --base-model "$RUNNER_TEMP/base/Human-GEM.yml" \
             --base-matrix "$RUNNER_TEMP/base/gene-essential.csv" \
             --base-label "$BASE_REF" \
-            --csv-url "$CSV_URL" \
             --summary-url "$SUMMARY_URL" \
             --out-summary "$RUNNER_TEMP/gene-essential-diff.md" \
             --out-detail "$RUNNER_TEMP/gene-essential-diff-detail.md" \

code/test/buildReport.py

@@ -45,8 +45,10 @@
 BASE_DIR = os.environ.get("BASE_RESULTS_DIR", "")
 BASE_REF = os.environ.get("BASE_REF", "the target branch")
 COMMIT_SHA = os.environ.get("COMMIT_SHA", "")
-# e.g. https://github.com/OWNER/REPO/blob/<branch>/data/testResults - used to link
-# each finding count to its CSV. Empty when run locally (then counts are plain text).
+# e.g. https://github.com/OWNER/REPO/blob/<branch>/data/testResults - used by the posted
+# comment to link each finding count to its CSV (relative links do not resolve in a
+# comment). The full-detail file uses relative links. Empty when run locally (then the
+# comment's counts are plain text).
 URL_BASE = os.environ.get("RESULTS_URL_BASE", "").rstrip("/")
 
 # Groups whose results are still being computed on this run; their rows show as
@@ -68,11 +70,11 @@ def _slug(label: str) -> str:
 
 
 def _labelled(label: str) -> str:
-    """The test name, linked to its explanation in the testResults README when the
-    repo URL is known (in CI); plain text when run locally. Full-detail file only --
-    the condensed comment links whole sections to the full file instead (see module
-    docstring)."""
-    return f"[{label}]({URL_BASE}/README.md#{_slug(label)})" if URL_BASE else label
+    """The test name, linked to its explanation in the testResults README. Full-detail
+    file only, which sits next to the README, so the link is relative and does not
+    depend on the branch; the condensed comment links whole sections to the full file
+    instead (see module docstring)."""
+    return f"[{label}](README.md#{_slug(label)})"
 
 
 # (label, key, kind, group, detail_file)
@@ -276,14 +278,17 @@ def _icon(value, base, kind):
     return "0", ":white_check_mark:", False, False
 
 
-def _cell(value, kind, detail, *, bold: bool = False) -> str:
+def _cell(value, kind, detail, *, bold: bool = False, relative: bool = False) -> str:
+    """The value, linked to its detail CSV when it is a positive count or a growth
+    failure. relative=True links next to the file (the full-detail file, which sits in
+    the same folder as the CSVs); otherwise the absolute URL is used, as the posted
+    comment needs, and the value stays plain when the repo URL is not known."""
     text = f"{value:.3g}" if kind == "growth" else (f"{value:.1f}" if kind == "score" else str(int(value)))
     if bold:
         text = f"**{text}**"
-    # link a positive count (or a growth failure) to its CSV, if we know the repo URL
     linkable = (kind == "count" and value) or (kind == "growth" and value <= 1e-6)
-    if URL_BASE and detail and linkable:
-        return f"[{text}]({URL_BASE}/{detail})"
+    if detail and linkable and (relative or URL_BASE):
+        return f"[{text}]({detail if relative else f'{URL_BASE}/{detail}'})"
     return text
 
 
@@ -314,7 +319,7 @@ def _full_table(computed: list[dict]) -> list[str]:
         if r["pending"]:
             lines.append(f"| {_labelled(r['label'])} | _running_ | | :hourglass_flowing_sand: |")
         else:
-            lines.append(f"| {_labelled(r['label'])} | {_cell(r['value'], r['kind'], r['detail'])} | "
+            lines.append(f"| {_labelled(r['label'])} | {_cell(r['value'], r['kind'], r['detail'], relative=True)} | "
                           f"{r['delta']} | {r['icon']} |")
     return lines
 
@@ -573,8 +578,7 @@ def main() -> int:
         "## Model quality report -- full detail",
         "",
         "_This is the full, per-check breakdown behind the pull-request comment's summary table._ "
-        + (f"_Row names link to their explanation in the [testResults README]({URL_BASE}/README.md)._"
-           if URL_BASE else ""),
+        "_Row names link to their explanation in the [testResults README](README.md)._",
         "",
         "### Model & network checks",
         "_Duplicate keys (model unloadable) and no growth block the merge; every other row "

code/test/diffGeneEssentiality.py

@@ -45,7 +45,7 @@
         --base-model /path/to/base/Human-GEM.yml \\
         --base-matrix /path/to/base/gene-essential.csv \\
         [--head-model model/Human-GEM.yml] [--head-matrix data/testResults/gene-essential.csv] \\
-        [--growth-tolerance 0.01] [--csv-url URL] [--summary-url URL] \\
+        [--growth-tolerance 0.01] [--summary-url URL] \\
         [--out-summary PATH] [--out-detail PATH] [--out-csv PATH]
 
 The base model/matrix are typically pulled from the target branch with ``git show``,
@@ -308,7 +308,6 @@ def build_report(
     *,
     growth_tolerance: float = 0.01,
     base_label: str = "the target branch",
-    csv_url: str = "",
     summary_url: str = "",
 ) -> tuple[str, str, str]:
     """Returns ``(comment_text, detail_text, detail_csv_text)``.
@@ -317,9 +316,9 @@ def build_report(
     request comment: icons and counts only, no gene names. ``detail_text`` is the
     per-gene breakdown behind it (which genes, what changed), meant to be appended to
     the committed ``gene-essential_summary.md`` rather than posted in the comment --
-    ``comment_text`` links to it instead of duplicating it. ``csv_url`` is the
-    committed blob URL of the per-gene, per-line CSV this call is about to write (e.g.
-    from ``--out-csv``), linked from ``detail_text``. ``summary_url`` is the committed
+    ``comment_text`` links to it instead of duplicating it. ``detail_text`` links to the
+    per-gene, per-line CSV (``--out-csv``) by file name, since both sit in
+    data/testResults; ``summary_url`` is the committed
     blob URL of ``gene-essential_summary.md`` (with ``detail_text`` appended to it),
     linked from ``comment_text``.
     """
@@ -455,7 +454,7 @@ def _neutral_status(rows: list[dict]) -> str:
     summary = "\n".join(summary_lines) + "\n"
 
     # --- detail: per-gene breakdown, appended to the committed gene-essential_summary.md ---
-    csv_ref = f"[gene-essential-diff.csv]({csv_url})" if csv_url else "the per-gene detail CSV"
+    csv_ref = "[gene-essential-diff.csv](gene-essential-diff.csv)"
     detail_lines = [
         f"### Gene essentiality: effect of this change (vs `{base_label}`)",
         "",
@@ -517,7 +516,6 @@ def main(argv: list[str] | None = None) -> int:
     parser.add_argument("--head-matrix", type=Path, default=DEFAULT_HEAD_MATRIX, help="gene-essential.csv for this branch (default: data/testResults/gene-essential.csv)")
     parser.add_argument("--growth-tolerance", type=float, default=0.01, help="minimum |growth ratio delta| to flag independent of task-based flips (default: 0.01)")
     parser.add_argument("--base-label", type=str, default="the target branch", help="name shown for the base branch, e.g. 'develop' (default: 'the target branch')")
-    parser.add_argument("--csv-url", type=str, default="", help="committed blob URL of --out-csv, linked from --out-detail instead of duplicating its content there")
     parser.add_argument("--summary-url", type=str, default="", help="committed blob URL of gene-essential_summary.md (with --out-detail appended to it), linked from --out-summary")
     parser.add_argument("--out-summary", type=Path, default=None, help="write the condensed, pull-request-comment text here instead of only stdout")
     parser.add_argument("--out-detail", type=Path, default=None, help="write the per-gene detail text here (meant to be appended to gene-essential_summary.md)")
@@ -527,7 +525,7 @@ def main(argv: list[str] | None = None) -> int:
     summary, detail_text, detail_csv = build_report(
         args.base_model, args.head_model, args.base_matrix, args.head_matrix,
         growth_tolerance=args.growth_tolerance, base_label=args.base_label,
-        csv_url=args.csv_url, summary_url=args.summary_url,
+        summary_url=args.summary_url,
     )
     print(summary)
     if args.out_summary:

data/deprecatedIdentifiers/deprecatedMetabolites.tsv

@@ -2395,3 +2395,10 @@ MAM01182m	MAM01182	xol7aone	C05455	HMDB0001993	CHEBI:17899	123743	LMST04030123
 MAM00592m	MAM00592	20ahchsterol	C05500	HMDB0006283	CHEBI:1296	440711	LMST01010201			20ahchsterol	MNXM1364606		m00592m
 MAM02754r	MAM02754	4nph	C00870	HMDB0001232	CHEBI:16836	980					MNXM526		
 MAM01019r	MAM01019	24nph	C02235	HMDB0002916	CHEBI:16318						MNXM1982		
+MAM03586c	MAM03586	galgluside_hs	C01290		CHEBI:17950		LMSP0501AB00			galgluside_hs	MNXM1108011		galgluside_hs_c
+MAM03586e	MAM03586	galgluside_hs	C01290		CHEBI:17950		LMSP0501AB00			galgluside_hs	MNXM1108011		galgluside_hs_s
+MAM03586g	MAM03586	galgluside_hs	C01290		CHEBI:17950		LMSP0501AB00			galgluside_hs	MNXM1108011		galgluside_hs_g
+MAM03586l	MAM03586	galgluside_hs	C01290		CHEBI:17950		LMSP0501AB00			galgluside_hs	MNXM1108011		galgluside_hs_l
+MAM01999c	MAM01999									fuc14galacglcgalgluside_hs	MNXM8835	m01999c	m01999c
+MAM01999e	MAM01999									fuc14galacglcgalgluside_hs	MNXM8835	m01999s	m01999s
+MAM01999g	MAM01999									fuc14galacglcgalgluside_hs	MNXM8835	m01999g	m01999g

data/deprecatedIdentifiers/deprecatedReactions.tsv

@@ -533,3 +533,15 @@ MAR20031							MNXR102281				0	RHEA:26205
 MAR00117		C4STMO2Pr				C4STMO2Pr	MNXR96419				0			C4STMO2Pr
 MAR00118		C4STMO2r				C4STMO2r	MNXR96420				0			C4STMO2r
 MAR02302		r1146				r1146					0			r1146
+MAR01348		GALGLUSIDEtg				GALGLUSIDEtg					0			GALGLUSIDEtg
+MAR01349		GALGLUSIDEtl				GALGLUSIDEtl					0			GALGLUSIDEtl
+MAR01351		GALGT1				GALGT1					0			GALGT1
+MAR12053		GLA2l				GLA2l					0			GLA2l
+MAR12065		NEU11l				NEU11l					0			NEU11l
+MAR12132		EX_galgluside_hs_e				EX_galgluside_hs[e]					0			EX_galgluside_hs[e]
+MAR01118		FUT34g				FUT34g	MNXR99739				0			FUT34g
+MAR01078		FUCFUCGALACGLCGALGLUSIDEtg				FUCFUCGALACGLCGALGLUSIDEtg	MNXR99694				0			FUCFUCGALACGLCGALGLUSIDEtg
+MAR01076		FUCFUCGALACGLCGALGLUSIDEte				FUCFUCGALACGLCGALGLUSIDEte	MNXR99694				0			FUCFUCGALACGLCGALGLUSIDEte
+MAR00595		EX_fucfucgalacglcgalgluside_hs_e				EX_fucfucgalacglcgalgluside_hs[e]					0			EX_fucfucgalacglcgalgluside_hs[e]
+MAR01141		FUT93g				FUT93g	MNXR99751				0			FUT93g
+MAR00766				r1410				HMR_0766			0			HMR_0766

data/testResults/README.md

@@ -21,12 +21,12 @@ own files. The pull request in each row is the one whose run last wrote those fi
 
 | Result file(s) | Produced by | Last updated by |
 | --- | --- | --- |
-| `qc_duplicate_keys.csv`, `qc_empty_reactions.csv`, `qc_annotation_consistency.csv`, `qc_deprecation_completeness.csv`, `qc_metabolite_completeness.csv`, `qc_reaction_sanity.csv`, `qc_duplicate_reactions.csv`, `qc_unused_entities.csv`, `qc_growth_blockers.csv` | `qcModelChecks.py` | **PR #1106** (model QC checks) |
-| `qc_annotation_issues.csv` | `annotationTest.py` | **PR #1106** (model QC checks) |
-| `qc_status.tsv` (round-trip, YAML lint, metabolic tasks, growth) | `testYamlConversion.py`, `testMetabolicTasks.py`, `action-yamllint`, `qcModelChecks.py` (via `qcStatus.py`) | **PR #1106** (model QC checks) |
-| `macaw_results.tsv`, `balance_results.csv`, `qc_structure_consistency.csv` | `macawTests.py`, `balanceTest.py`, `structureConsistencyTest.py` | **PR #1106** (MACAW and balance) |
-| `memote_score.md` | `memoteSnapshot.py` (fast subset every PR; full suite via `/run memote`) | **PR #1106** (MEMOTE) |
-| `gene-essential.csv`, `gene-essential_summary.md` | `gradedEssentiality.py` via `/run gene-essentiality` | **PR #1106** (gene essentiality) |
+| `qc_duplicate_keys.csv`, `qc_empty_reactions.csv`, `qc_annotation_consistency.csv`, `qc_deprecation_completeness.csv`, `qc_metabolite_completeness.csv`, `qc_reaction_sanity.csv`, `qc_duplicate_reactions.csv`, `qc_unused_entities.csv`, `qc_growth_blockers.csv` | `qcModelChecks.py` | **PR #1107** (model QC checks) |
+| `qc_annotation_issues.csv` | `annotationTest.py` | **PR #1107** (model QC checks) |
+| `qc_status.tsv` (round-trip, YAML lint, metabolic tasks, growth) | `testYamlConversion.py`, `testMetabolicTasks.py`, `action-yamllint`, `qcModelChecks.py` (via `qcStatus.py`) | **PR #1107** (model QC checks) |
+| `macaw_results.tsv`, `balance_results.csv`, `qc_structure_consistency.csv` | `macawTests.py`, `balanceTest.py`, `structureConsistencyTest.py` | **PR #1107** (MACAW and balance) |
+| `memote_score.md` | `memoteSnapshot.py` (fast subset every PR; full suite via `/run memote`) | **PR #1107** (MEMOTE) |
+| `gene-essential.csv`, `gene-essential_summary.md` | `gradedEssentiality.py` via `/run gene-essentiality` | **PR #1107** (gene essentiality) |
 
 ## 2. What each check means
 

data/testResults/balance_results.csv

@@ -218,9 +218,6 @@ MAR10036,acyl-CoA pool formation and breakdown,C:16.9524;H:31.9518;R:-1,
 MAR10037,cholesterol-ester pool formation and breakdown,C:16.9524;H:31.9518;R:-1,
 MAR10038,cholesterol-ester pool formation and breakdown,C:16.9524;H:31.9518;R:-1,
 MAR10061,fatty acyl-CoA reduction,C:-1;H:-2,
-MAR12991,3HPVSCOAhc,,-1
-MAR12992,3HPVSTEThc,,1
-MAR12994,FVSCOAhc,,-1
 MAR12995,MDZGLChr,,-1
 MAR13078,PROTEIN_BS,C:-32.5694;H:-66.0418;N:-9.3852;O:-16.5863;S:-0.283;X:1,-0.202
 MAR13081,CYOOm3i,,0.06