@datagrok/proteomics 1.2.0 → 1.3.0

package/docs/personas-and-capabilities.md

@@ -15,6 +15,12 @@ boundary to be real, not conventional.
 | **Proteomics analyst** (data generator) | Builds the analysis: import → annotate → normalize → impute → differential expression | A defensible, reproducible, FDR-controlled result they can stand behind and hand off |
 | **Biology scientist** (data consumer / reviewer) | Interprets a finished analysis: volcano, enrichment, per-protein annotation | To understand *what changed and what it means* — without touching (or being able to change) the analysis itself |
 
+A third role sits **outside** this pair: the **package administrator** (deployment owner)
+configures the Share-for-review governance — the controlled list of projects analysts may
+share under, the default reviewer team, and the review Spaces (see
+[Configuration](../README.md#configuration)). They neither build nor interpret analyses; they
+set, once per deployment, *where* shared results land and *under what project names*.
+
 The dividing line is **"the interpretation"** — the set of decisions that determine what the
 result *says*: group assignment, normalization, imputation, the DE method, and the
 significance thresholds. The analyst owns these. The biologist reads the consequences and
@@ -85,6 +91,13 @@ group has any Edit / Share / Delete right at the project, child-space, or umbrel
 level. So through the **shared snapshot**, a biologist genuinely cannot change the
 interpretation — they get a read-only copy.
 
+The dialog itself is **governed by package settings the administrator maintains** (Plugins →
+Proteomics → Settings). The analyst files the snapshot under a **project** chosen from an
+admin-curated controlled vocabulary — they cannot invent project names, and cannot share at
+all until an admin has populated the list — and it goes to a reviewer **team** whose default
+the admin can preset. This keeps shared analyses landing under approved project names, in the
+intended teams and Spaces, rather than letting each analyst decide that governance ad hoc.
+
 ## The gap
 
 Enforcement today is **data-shape + access**, not **identity**. Concretely:
@@ -163,3 +176,6 @@ an explicit read-only "Reviewer mode" toggle vs. inferring it from group members
 - `src/menu.ts` — `sampleLevelDisabledReason`, the `sampleOnly` gate
 - `src/package.ts` — `requireSampleLevelData`, the handler-level guards
 - `src/publishing/` — the Share for Review view-only publish flow
+- `src/publishing/package-settings-editor.ts` — the administrator's settings editor
+  (project vocabulary, default reviewer team)
+- [`README.md` → Configuration](../README.md#configuration) — the admin-configurable package settings

package/docs/publishing-design-rationale.md

@@ -0,0 +1,140 @@
+# Publishing (Share for Review) — design rationale
+
+Why the Share-for-review subsystem is built the way it is, and **which parts are permanent
+architecture vs. workaround weight that the base platform could absorb.**
+
+This is the reference behind the [personas & capabilities](./personas-and-capabilities.md) doc's
+enforcement claims. That doc says *what* the read-only boundary is; this says *why the code takes
+the shape it does to make that boundary real* — and where we'd rather the platform did the work.
+
+Audience: anyone questioning the complexity of `src/publishing/`, and anyone planning what
+platform capabilities to grow so this package can shed code. Per our standing preference, a
+workaround defended as *temporary, with a named platform successor* is a stronger position than
+one defended as permanent design.
+
+## The one-sentence defense
+
+Most of `src/publishing/` (~1,900 lines across 9 files) is not "sharing." It is the machinery to
+**guarantee three properties** a compliance-sensitive review needs, none of which the platform
+gave us for free at build time:
+
+1. **Provably read-only** — the reviewer group *cannot* alter the interpretation.
+2. **Provably minimized** — only the interpretation ships; raw/again-identifying data does not.
+3. **Provably reload-survivable + versioned** — what the reviewer opens is what was published,
+   every version is immutable, and supersession is auditable.
+
+The naming scheme and the use of Spaces (the two most visible choices) are *downstream* of these
+guarantees, not the point of them.
+
+## Design choices, grouped by the guarantee they serve
+
+Legend for **Verdict**: 🟢 keep in package (genuinely Proteomics-specific) · 🟡 push to platform
+(generic capability we had to hand-roll) · 🔵 hybrid (platform does the mechanism, package keeps the policy)
+
+### A. Read-only is enforced, not assumed
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| A1 | Grant **View directly on the Project**, not via Space inheritance | `publish-project.ts:323` | Spike A2: `permissions.get(project)` does **not** surface Space-inherited grants, so an inherited grant is unverifiable. | 🟡 |
+| A2 | **Verify-and-rollback ACL gate** — after granting, re-read permissions across three rings (project, child Space, umbrella Space); any `edit`/`share`/`delete` for the reviewer group ⇒ delete the project and abort | `publish-project.ts:331-370` | The feature refuses to ship what it cannot prove is read-only. This is the single most important — and most expensive — thing the code does. | 🔵 |
+| A3 | Share to a **group/team, never individuals**; eligible groups exclude hidden, personal, and "All users" | `reviewer-groups.ts:9-19` | Review is a team activity; "All users" would defeat the point of a scoped share. | 🔵 |
+
+### B. Data minimization — a trimmed snapshot, not a live link
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| B1 | **Deep clone; source never mutated** — the publish path operates only on the clone | `trim-dataframe.ts:75-120` | Reviewers get a frozen point-in-time copy; later edits to the working analysis cannot leak. | 🟢 |
+| B2 | **Column allowlist** (Protein ID, Gene, log2FC, p, adj.p, significant, [direction, quantities]); everything else dropped | `trim-dataframe.ts:108-120` | The generalized form of the 1.2.0 "client-name removal." A *whitelist* means a new, unexpected column is excluded by default — safer than a blocklist. | 🟢 |
+
+### C. Provenance is embedded and redundant
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| C1 | Rich metadata baked into the snapshot (who/email/DE method/thresholds/version/UUID/date) | `publish-state.ts:71-92`, `trim-dataframe.ts:158-174` | The snapshot is self-describing — provenance travels with the data, not a side channel. | 🟢 |
+| C2 | **Dual-write belt-and-braces** — every field written as *both* a DataFrame tag *and* a hidden `_meta_*` column; read column-first, tag-second | `publish-state.ts:24-60,137-225` | Defends against serialization paths that strip tags but keep columns (or vice versa) — an observed failure mode. | 🟡 |
+| C3 | Numeric metadata stored **as strings** | `trim-dataframe.ts:153-167` | `addNewFloat` is Float32; string storage preserves exact double thresholds. | 🔵 |
+
+### D. Versioning is first-class and non-destructive
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| D1 | **Opinionated naming** `Proteomics-Review-<slug>-v<N>-<date>`; structural suffix fixed (not configurable) | `publish-project.ts:244-250`, `publish-settings.ts` | The `-v<N>-` segment is parsed to detect prior versions; letting it vary would break version detection. Prefix/Space *are* configurable. | 🔵 |
+| D2 | **Republish detection** by name pattern → bump to v(N+1); prior version stays available (never overwritten) | `publish-state.ts:279-328`, `publish-project.ts:138-140` | Immutable, auditable version history. | 🔵 |
+| D3 | **Bidirectional supersede pointers** (`supersedes`/`supersededBy`) on both projects, dual-written to project options *and* the DataFrame | `publish-project.ts:436-478` | Either version can be navigated to the other; survives whichever persistence path holds. | 🟡 |
+
+### E. Round-trip fidelity is verified
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| E1 | **Reopen-and-assert gate** (default on) — `closeAll` → reopen → assert published shape survives a reload; rollback on failure | `publish-project.ts:372-434`, `assert-published-shape.ts` | Proves the reviewer sees exactly what was published, before telling the analyst it worked. | 🟡 |
+| E2 | **Formula-line self-heal** — a serializer strips volcano threshold lines; re-apply from tags and re-check once before rollback | `publish-project.ts:401-433`, `post-open-recovery.ts` | Works around look-config loss on project round-trip (Phase 13 evidence). | 🟡 |
+| E3 | **Re-render viewers on the trimmed snapshot** so the shared dashboard is self-contained | `publish-project.ts:265-295` | The published project must stand alone — no dependency on the analyst's session tables. | 🔵 |
+| E4 | **Post-open recovery autostart** — re-applies formula lines and re-wires enrichment↔volcano cross-highlight on reopen | `post-open-recovery.ts`, `package.ts:1010-1037` | Interaction subscriptions and look config don't survive serialization; rebuild them on open. | 🟡 |
+
+### F. Where content lands + idempotency
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| F1 | Publish to **named Spaces, not a personal workspace** | `publish-project.ts:178-239` | A shared, governed location the reviewer team can browse — not tied to one user's private space. | 🔵 |
+| F2 | **Two-level Space hierarchy** — umbrella `Proteomics-Reviews` + per-project child Space | `publish-project.ts:178-239` | Groups all reviews under one root, one child per project — a navigable tree instead of a flat pile. | 🔵 |
+| F3 | **Create-first-then-enumerate idempotency** for Spaces; fall back to `spaces.list()` on "already exists" | `publish-project.ts:184-238` | The smart-filter API does not reliably surface Space-flagged Projects, so enumeration is the reliable lookup. | 🟡 |
+
+### G. Governance + workflow UX
+
+| # | Choice | Where | Why | Verdict |
+|---|---|---|---|---|
+| G1 | **Controlled project vocabulary** — analyst picks from an admin-maintained list, never free-types | `share-dialog.ts:57-73`, `publish-settings.ts` | Keeps shared results filed under sanctioned project names; a governance control. | 🟢 |
+| G2 | **Precondition gate** — share only available after DE completes | `share-dialog.ts:31-34` | There is no interpretation to review until DE has run. | 🟢 |
+| G3 | **Reactive confirmation summary** — exact resulting name/team/supersede status shown before OK | `share-dialog.ts:109-151` | No surprises; the analyst sees the outcome before committing. | 🟢 |
+| G4 | **Workspace restoration** after publish — tear down verification artifacts, restore the analyst's original view | `publish-project.ts:480-506` | The verify/supersede steps `closeAll` repeatedly; put the analyst back where they were. | 🔵 |
+| G5 | **Full package-settings configurability** (Space, prefix, default team, vocabulary, verify default) | `publish-settings.ts` | Retarget a deployment without a code change. | 🟢 |
+
+## Mapping to platform capabilities
+
+Reading the 🟡/🔵 rows top-down, the workaround weight collapses to a small number of platform
+gaps. If the platform closed these, the package could delete or thin most of `src/publishing/`.
+
+| Platform gap (observed) | Evidence in code | What a platform capability would give us | Package code it retires |
+|---|---|---|---|
+| **P1. No first-class "publish read-only snapshot to a group" primitive** | The entire orchestrator exists to compose this from lower-level parts | A single `publishReadOnly(entity, group, {snapshot, immutable})` that grants view-only, freezes, and returns a stable handle | Most of `publish-project.ts` (A1, A2, F1–F3) |
+| **P2. `permissions.get` doesn't reflect Space-inherited grants; may not report share/delete** | Spike A1/A2 notes at `publish-project.ts:113-118,331-370` | A permissions read that returns the *effective* ACL (own + inherited) across all verbs | A2 three-ring check collapses to one `get`; A1 direct-grant workaround unneeded |
+| **P3. Serialization strips look config (formula lines) and interaction subscriptions** | E2/E4; `post-open-recovery.ts`, `assert-published-shape.ts:11-17` | Viewer look + formula lines + declared subscriptions persist across `save→open` | E2, E4 self-heal and recovery hook; part of E1's assertions |
+| **P4. Tag map / metadata not reliably durable across `clone`/`save`** | C2 dual-write; `publish-state.ts:227-255`, `trim-dataframe.ts:20` | Durable, typed entity/DataFrame metadata that survives clone and round-trip | C2 `_meta_*` columns; the column-first/tag-second reader in `getPublishedMetadata` |
+| **P5. Smart-filter can't reliably find Space-flagged Projects** | F3 note at `publish-project.ts:184-203` | `dapi.projects.filter(isSpace=…).first()` that actually works | F3 create-first-then-enumerate dance |
+| **P6. No immutable/versioned entity semantics** | D1–D3 name-parsing + manual supersede pointers | Native entity versioning with supersede links | D2/D3 name-pattern versioning and dual-written pointers; D1's suffix could relax |
+| **P7. No round-trip integrity check** | E1; `assert-published-shape.ts` | If P3+P4 land, publish *is* faithful by construction and E1 becomes optional | E1 gate downgrades to a test-only assertion |
+
+### What stays in the package regardless
+
+The 🟢 rows are the genuine Proteomics policy and don't belong in the platform:
+
+- **B1/B2** — *which* columns constitute "the interpretation" (the allowlist) is domain knowledge.
+- **C1** — *which* provenance fields matter (DE method, FC/p thresholds) is domain knowledge.
+- **G1/G2/G3/G5** — the vocabulary, the DE precondition, the confirmation copy, the settings surface.
+
+Even under a fully-featured platform, the package would still call a `publishReadOnly` primitive,
+hand it a domain-trimmed snapshot with domain provenance, and gate the UI on DE completion. That
+residue is small — roughly `trim-dataframe.ts` + `share-dialog.ts` + `publish-settings.ts`.
+
+## Recommended platform asks, prioritized
+
+1. **P1 + P2 together** — the read-only-to-group primitive with a trustworthy effective-permissions
+   read. Highest leverage: retires the bulk of the orchestrator *and* the code we're least confident
+   in (the ACL gate can only be as correct as `permissions.get`).
+2. **P3 + P4** — durable look/subscription/metadata across round-trip. Retires the self-heal and
+   recovery machinery, which is the most brittle code in the subsystem.
+3. **P6** — native immutable versioning. Retires the name-parsing version scheme and lets the naming
+   suffix relax.
+4. **P5, P7** — fall out naturally once the above land.
+
+## Honest caveats to raise proactively
+
+- **The ACL gate is only as strong as `permissions.get`.** If P2 is real (the API may not report
+  `share`/`delete`), the three-ring check at `publish-project.ts:358-361` cannot detect those leaks
+  today. Worth verifying against the current platform release before claiming the guarantee is
+  airtight. *(This claim is inferred from the A1 spike note and should be re-checked live.)*
+- **Several verify/rollback steps `closeAll` the user's session.** G4 restoration mitigates it, but
+  it's inherent to doing round-trip verification in the same client session; P7 landing would remove
+  the need entirely.
+- **This is workaround-heavy by design, and that's defensible** *because* each workaround has a named
+  platform successor above — not because the complexity is intrinsic to sharing.

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@datagrok/proteomics",
   "friendlyName": "Proteomics",
-  "version": "1.2.0",
+  "version": "1.3.0",
   "author": {
     "name": "Ed Jaeger",
     "email": "ejaeger@datagrok.ai"
@@ -17,8 +17,8 @@
     "@datagrok-libraries/math": "^1.2.6",
     "@datagrok-libraries/ml": "^6.10.7",
     "@datagrok-libraries/statistics": "^1.2.12",
-    "@datagrok-libraries/utils": "^4.6.9",
     "@datagrok-libraries/test": "^1.1.0",
+    "@datagrok-libraries/utils": "^4.6.9",
     "cash-dom": "^8.1.5",
     "datagrok-api": "^1.25.0",
     "rxjs": "^6.5.5",
@@ -61,27 +61,42 @@
   "properties": [
     {
       "name": "reviewSpaceName",
-      "description": "Umbrella Space that shared review snapshots are published under. Each target gets a child Space inside it.",
+      "description": "Umbrella Space that shared review snapshots are published under. Each project gets a child Space inside it.",
       "propertyType": "string",
       "defaultValue": "Proteomics-Reviews",
       "nullable": false
     },
     {
       "name": "reviewNamePrefix",
-      "description": "Prefix for shared review project names. The structural '-<target>-v<version>-<date>' suffix stays fixed so re-share version detection keeps working.",
+      "description": "Prefix for shared review project names. The structural '-<project>-v<version>-<date>' suffix stays fixed so re-share version detection keeps working.",
       "propertyType": "string",
       "defaultValue": "Proteomics-Review",
       "nullable": false
     },
+    {
+      "name": "defaultReviewerGroup",
+      "description": "Team pre-selected in the Share for Review dialog's 'Share with team' picker. Leave blank to default to the first available team.",
+      "propertyType": "string",
+      "defaultValue": "",
+      "nullable": true
+    },
     {
       "name": "verifyPublishedDashboard",
       "description": "After publishing, re-open the shared dashboard to verify it survives a reload. Safer for client deliverables; turn off for faster demo shares.",
       "propertyType": "bool",
       "defaultValue": true,
       "nullable": false
+    },
+    {
+      "name": "projectVocabulary",
+      "description": "Comma-separated controlled list of project names an analyst can share under. Package administrators add, edit, or remove names here; the Share for Review dialog offers exactly these as a dropdown (no free text).",
+      "propertyType": "string",
+      "defaultValue": "",
+      "nullable": true
     }
   ],
   "overrides": {
-    "datagrok-api": "$datagrok-api"
+    "datagrok-api": "$datagrok-api",
+    "uuid": "^11.1.1"
   }
 }

package/src/analysis/imputation.ts

@@ -7,7 +7,9 @@ import {getGroups} from './experiment-setup';
 
 /** Generates a random number from a normal distribution using Box-Muller transform. */
 function randomNormal(mean: number, stdev: number): number {
-  const u1 = Math.random();
+  // Draw u1 in (0, 1] rather than [0, 1): Math.random() can return exactly 0,
+  // which would give log(0) = -Infinity and an Infinity/NaN draw.
+  const u1 = 1 - Math.random();
   const u2 = Math.random();
   const z = Math.sqrt(-2 * Math.log(u1)) * Math.cos(2 * Math.PI * u2);
   return mean + stdev * z;

package/src/analysis/normalization.ts

@@ -59,7 +59,10 @@ export function quantileNormalize(df: DG.DataFrame, colNames: string[]): void {
 
   // Find the maximum number of non-null values across columns
   const maxValid = Math.max(...colData.map((d) => d.indices.length));
-  if (maxValid === 0) return;
+  // Need at least 2 values in the most-populated column: the rank interpolation
+  // below divides by (maxValid - 1), and quantile normalization is undefined for
+  // a single quantile point.
+  if (maxValid < 2) return;
 
   // Compute rank means using scaled rank alignment
   // For each rank position r (0..maxValid-1), compute the mean of values
@@ -69,7 +72,7 @@ export function quantileNormalize(df: DG.DataFrame, colNames: string[]): void {
     let sum = 0;
     let count = 0;
     for (const d of colData) {
-      if (d.indices.length === 0) continue;
+      if (d.indices.length < 2) continue;
       // Map rank r in [0..maxValid-1] to position in this column's sorted values
       const pos = (r / (maxValid - 1)) * (d.indices.length - 1);
       // Interpolate between floor and ceil positions
@@ -86,10 +89,12 @@ export function quantileNormalize(df: DG.DataFrame, colNames: string[]): void {
   // Replace each column's sorted values with the corresponding rank mean
   for (const d of colData) {
     const n = d.indices.length;
-    if (n === 0) continue;
+    // n < 2: a single-value column has no rank spread; (r / (n - 1)) would be
+    // 0/0 = NaN. Leave its lone value unchanged rather than corrupting it.
+    if (n < 2) continue;
     for (let r = 0; r < n; r++) {
       // Map this column's rank to the global rank mean position
-      const globalPos = maxValid === 1 ? 0 : (r / (n - 1)) * (maxValid - 1);
+      const globalPos = (r / (n - 1)) * (maxValid - 1);
       const lo = Math.floor(globalPos);
       const hi = Math.min(Math.ceil(globalPos), maxValid - 1);
       const frac = globalPos - lo;

package/src/analysis/spc-storage.ts

@@ -111,7 +111,7 @@ function baselinePath(instrumentId: string, opts?: StorageOpts): string {
 }
 
 // =====================================================================================
-// Slugify — T-16-01 mitigation. Mirrors publishing/publish-state.ts:slugifyTarget but
+// Slugify — T-16-01 mitigation. Mirrors publishing/publish-state.ts:slugifyProject but
 // PRESERVES case so instrument names like 'QExactive-01' round-trip exactly.
 // =====================================================================================
 

package/src/menu.ts

@@ -40,10 +40,11 @@ export interface ProteomicsMenuHandlers {
   showVolcanoPlot: () => void;
   showHeatmap: () => void;
   showPcaPlot: () => void;
-  showGroupMeanCorrelation: () => void;
   showQcDashboard: () => void;
   showSpcDashboard: () => void;
   enrichmentCharts: () => void;
+  rankAbundance: () => void;
+  abundanceCorrelation: () => void;
   showAllVisualizations: () => void;
   shareAnalysisForReview: () => void;
 }
@@ -121,10 +122,12 @@ export function buildProteomicsRibbonMenu(view: DG.TableView, h: ProteomicsMenuH
   viz.item('Volcano Plot...', h.showVolcanoPlot);
   viz.item('Heatmap...', h.showHeatmap, null, sampleOnly);
   viz.item('PCA...', h.showPcaPlot, null, sampleOnly);
-  viz.item('Group-Mean Correlation...', h.showGroupMeanCorrelation, null, sampleOnly);
   viz.item('QC Dashboard...', h.showQcDashboard, null, sampleOnly);
   viz.item('SPC Dashboard...', h.showSpcDashboard);
   viz.item('Enrichment Charts...', h.enrichmentCharts);
+  // Not sampleOnly: both work on Candidates (AVG Group Quantity) and annotated Reports.
+  viz.item('Rank–Abundance', h.rankAbundance);
+  viz.item('Abundance Correlation', h.abundanceCorrelation);
   viz.item('Show All Visualizations...', h.showAllVisualizations, null, sampleOnly);
 
   const share = root.group('Share');

package/src/package-api.ts

@@ -95,8 +95,8 @@ export namespace funcs {
     return await grok.functions.call('Proteomics:ShowPcaPlot', {});
   }
 
-  export async function showGroupMeanCorrelation(): Promise<void> {
-    return await grok.functions.call('Proteomics:ShowGroupMeanCorrelation', {});
+  export async function abundanceCorrelation(): Promise<void> {
+    return await grok.functions.call('Proteomics:AbundanceCorrelation', {});
   }
 
   export async function showQcDashboard(): Promise<void> {
@@ -119,6 +119,10 @@ export namespace funcs {
     return await grok.functions.call('Proteomics:ExportEnrichmentInputs', {});
   }
 
+  export async function rankAbundance(): Promise<void> {
+    return await grok.functions.call('Proteomics:RankAbundance', {});
+  }
+
   export async function enrichmentCharts(): Promise<void> {
     return await grok.functions.call('Proteomics:EnrichmentCharts', {});
   }
@@ -153,6 +157,10 @@ export namespace funcs {
     return await grok.functions.call('Proteomics:PublishedAnalysisPanelWidget', { proteinId });
   }
 
+  export async function proteomicsSettingsEditor(propList: any ): Promise<any> {
+    return await grok.functions.call('Proteomics:ProteomicsSettingsEditor', { propList });
+  }
+
   export async function recoverPublishedProjectsOnStartup(): Promise<void> {
     return await grok.functions.call('Proteomics:RecoverPublishedProjectsOnStartup', {});
   }

package/src/package-test.ts

@@ -11,6 +11,8 @@ import './tests/enrichment-visualization';
 import './tests/spectronaut-parser';
 import './tests/spectronaut-candidates-parser';
 import './tests/spectronaut-candidates-e2e';
+import './tests/rank-abundance';
+import './tests/abundance-correlation';
 import './tests/fragpipe-parser';
 import './tests/fragpipe-e2e';
 import './tests/subcellular-location';
@@ -20,9 +22,9 @@ import './tests/volcano';
 import './tests/gene-label-resolver';
 import './tests/smart-pathway-filter';
 import './tests/uniprot-panel';
-import './tests/group-mean-correlation';
 import './tests/publish-spike';
 import './tests/publish-roundtrip';
+import './tests/project-vocabulary';
 import './tests/spc-formula-lines-spike';
 import './tests/spc';
 

package/src/package.g.ts

@@ -83,9 +83,9 @@ export async function showPcaPlot() : Promise<void> {
   await PackageFunctions.showPcaPlot();
 }
 
-//name: showGroupMeanCorrelation
-export async function showGroupMeanCorrelation() : Promise<void> {
-  await PackageFunctions.showGroupMeanCorrelation();
+//name: abundanceCorrelation
+export async function abundanceCorrelation() : Promise<void> {
+  await PackageFunctions.abundanceCorrelation();
 }
 
 //name: showQcDashboard
@@ -113,6 +113,11 @@ export async function exportEnrichmentInputs() : Promise<void> {
   await PackageFunctions.exportEnrichmentInputs();
 }
 
+//name: rankAbundance
+export async function rankAbundance() : Promise<void> {
+  await PackageFunctions.rankAbundance();
+}
+
 //name: enrichmentCharts
 export async function enrichmentCharts() : Promise<void> {
   await PackageFunctions.enrichmentCharts();
@@ -155,6 +160,15 @@ export function publishedAnalysisPanelWidget(proteinId: string) : any {
   return PackageFunctions.publishedAnalysisPanelWidget(proteinId);
 }
 
+//name: Proteomics package settings editor
+//tags: packageSettingsEditor
+//input: object propList 
+//output: widget result
+//meta.role: packageSettingsEditor
+export async function proteomicsSettingsEditor(propList: any) : Promise<any> {
+  return await PackageFunctions.proteomicsSettingsEditor(propList);
+}
+
 //tags: autostart
 //meta.autostartImmediate: true
 export async function recoverPublishedProjectsOnStartup() : Promise<void> {

package/src/package.ts

@@ -30,15 +30,17 @@ import {createExpressionHeatmap} from './viewers/heatmap';
 import {createPcaPlot} from './viewers/pca-plot';
 import {openQcDashboard} from './viewers/qc-dashboard';
 import {openSpcDashboard} from './viewers/spc-dashboard';
-import {createGroupMeanCorrelation} from './viewers/group-mean-correlation';
+import {openAbundanceCorrelation} from './viewers/abundance-correlation';
 import {uniprotPanel} from './panels/uniprot-panel';
 import {focusProtein} from './panels/protein-focus';
 import {publishedAnalysisPanel} from './panels/published-analysis-panel';
 import {showShareForReviewDialog} from './publishing/share-dialog';
+import {proteomicsSettingsEditorWidget} from './publishing/package-settings-editor';
 import {recoverPublishedProject} from './publishing/post-open-recovery';
 import {isPublished} from './publishing/publish-state';
 import {showEnrichmentDialog} from './analysis/enrichment';
 import {openEnrichmentVisualization} from './viewers/enrichment-viewers';
+import {openRankAbundance} from './viewers/rank-abundance';
 import {findColumn} from './utils/column-detection';
 import {SEMTYPE, DEFAULT_FC_THRESHOLD, DEFAULT_P_THRESHOLD} from './utils/proteomics-types';
 import {buildProteomicsRibbonMenu} from './menu';
@@ -341,10 +343,11 @@ export class PackageFunctions {
       showVolcanoPlot: () => PackageFunctions.showVolcanoPlot(),
       showHeatmap: () => PackageFunctions.showHeatmap(),
       showPcaPlot: () => PackageFunctions.showPcaPlot(),
-      showGroupMeanCorrelation: () => PackageFunctions.showGroupMeanCorrelation(),
+      abundanceCorrelation: () => PackageFunctions.abundanceCorrelation(),
       showQcDashboard: () => PackageFunctions.showQcDashboard(),
       showSpcDashboard: () => PackageFunctions.showSpcDashboard(),
       enrichmentCharts: () => PackageFunctions.enrichmentCharts(),
+      rankAbundance: () => PackageFunctions.rankAbundance(),
       showAllVisualizations: () => PackageFunctions.showAllVisualizations(),
       shareAnalysisForReview: () => PackageFunctions.shareAnalysisForReview(),
     };
@@ -824,19 +827,12 @@ export class PackageFunctions {
   }
 
   @grok.decorators.func()
-  static async showGroupMeanCorrelation(): Promise<void> {
-    const tv = grok.shell.tv;
-    const df = tv?.dataFrame;
-    if (!tv || !df) { grok.shell.warning('No table open'); return; }
-    if (!requireDifferentialExpression(df,
-      'Run Differential Expression first (Proteomics | Analyze | Differential Expression)')) return;
-    const groups = getGroups(df);
-    if (!groups) {
-      grok.shell.warning('Annotate experimental groups first (Proteomics | Annotate Experiment)');
-      return;
-    }
-    const sp = createGroupMeanCorrelation(df);
-    tv.addViewer(sp);
+  static async abundanceCorrelation(): Promise<void> {
+    const df = grok.shell.tv?.dataFrame;
+    if (!df) { grok.shell.warning('No table open'); return; }
+    // Works on any analyzed protein table (Candidates or annotated Report) —
+    // openAbundanceCorrelation itself warns when no per-condition abundance is present.
+    openAbundanceCorrelation(df);
   }
 
   @grok.decorators.func()
@@ -898,6 +894,15 @@ export class PackageFunctions {
     showEnrichmentInputExportDialog(df);
   }
 
+  @grok.decorators.func()
+  static async rankAbundance(): Promise<void> {
+    const df = grok.shell.tv?.dataFrame;
+    if (!df) { grok.shell.warning('No table open'); return; }
+    // Works on any analyzed protein table (Candidates or annotated Report) —
+    // openRankAbundance itself warns when no per-condition abundance is present.
+    openRankAbundance(df);
+  }
+
   @grok.decorators.func()
   static async enrichmentCharts(): Promise<void> {
     const tv = grok.shell.tv;
@@ -990,6 +995,17 @@ export class PackageFunctions {
     return publishedAnalysisPanel(proteinId);
   }
 
+  @grok.decorators.func({
+    name: 'Proteomics package settings editor',
+    meta: {role: 'packageSettingsEditor'},
+    tags: ['packageSettingsEditor'],
+  })
+  static async proteomicsSettingsEditor(
+    @grok.decorators.param({name: 'propList', type: 'object'}) propList: DG.Property[],
+  ): Promise<DG.Widget> {
+    return proteomicsSettingsEditorWidget(propList);
+  }
+
   @grok.decorators.func({tags: ['autostart'], meta: {autostartImmediate: 'true'}})
   static async recoverPublishedProjectsOnStartup(): Promise<void> {
     const tryEvent: any =

package/src/panels/published-analysis-panel.ts

@@ -12,7 +12,7 @@ import {getGroups} from '../analysis/experiment-setup';
 /**
  * Reviewer-side audit context panel. Opens when a reviewer clicks a protein
  * row in a published Project. Surfaces the metadata the analyst stamped at
- * publish time (DE method, thresholds, group names, target, share date, sharer
+ * publish time (DE method, thresholds, group names, project, share date, sharer
  * name) plus PUB-13 mailto button. First-line `isPublished(df)` guard ensures
  * it does NOT render on the analyst's live working DataFrame.
  *
@@ -123,7 +123,7 @@ export function publishedAnalysisPanel(proteinId: string): DG.Widget {
     ? `${meta.pThreshold} (FDR-adjusted)`
     : '(unknown)';
 
-  body.appendChild(fieldRow('Target', meta.target || '(unknown)'));
+  body.appendChild(fieldRow('Project', meta.project || '(unknown)'));
   body.appendChild(fieldRow('Shared', dateSlice(meta.publishedAt)));
   body.appendChild(fieldRow('Shared by', meta.publishedBy || '(unknown)'));
   body.appendChild(fieldRow('Method', methodLabel(meta.deMethod)));

package/src/publishing/assert-published-shape.ts

@@ -127,8 +127,8 @@ export async function assertPublishedShape(
   }
   const expectedMeta = contract.expectedMeta;
   const eqFloat = (a: number, b: number) => !(Math.abs(a - b) > 1e-9);
-  if (meta.target !== expectedMeta.target)
-    throw new Error(`assertPublishedShape: meta.target: got "${meta.target}", expected "${expectedMeta.target}"`);
+  if (meta.project !== expectedMeta.project)
+    throw new Error(`assertPublishedShape: meta.project: got "${meta.project}", expected "${expectedMeta.project}"`);
   if (meta.deMethod !== expectedMeta.deMethod)
     throw new Error(`assertPublishedShape: meta.deMethod: got "${meta.deMethod}", expected "${expectedMeta.deMethod}"`);
   if (!eqFloat(meta.fcThreshold, expectedMeta.fcThreshold))