---
vault_clearance: KETER
halo:
  classification: RESTRICTED
  confidence: HIGH
  front: "39_Project_TheUnreleased — bounty board"
  custodian: "Jixiang Leng"
  created: 2026-05-04
  wing: READY
  containment: "Operator-internal experimental program. Each bounty has a verification gate; SOLVED requires the gate to pass + WORLDLINE entry."
---

# Project 39 — Bounty Board

Time-boxed experiments to test the load-vs-release hypothesis from [HALO_LOAD_VS_RELEASE.md](HALO_LOAD_VS_RELEASE.md). Each bounty corresponds to one of the 6 testable predictions or supports the experimental infrastructure.

**Status legend:** `OPEN` · `IN PROGRESS` · `BLOCKED` · `SOLVED` · `DEFERRED` · `FALSIFIED` (prediction failed; logged as honest negative).

**Bounty ID prefix:** `UR-` for The Unreleased.

---

## Tier 0 — Already done (the discovery itself)

| ID | Title | Status | Receipts |
|---|---|---|---|
| **UR-D0a** | D03 Calandrelli HUVEC peak-and-release replication (proliferative-capable cell baseline) | **DONE 2026-05-04** | [`38_Project_Firewall/replications/calandrelli_d03_huvec/MEMO_2026-05-04_FULL_REPLICATION.md`](../38_Project_Firewall/replications/calandrelli_d03_huvec/MEMO_2026-05-04_FULL_REPLICATION.md) — all 4 criteria PASSED |
| **UR-D0b** | Rotenone DA neuron monotone-load observation (post-mitotic cell baseline) | **DONE 2026-05-04** | [`38_Project_Firewall/replications/rotenone_DA_neurons/MEMO.md`](../38_Project_Firewall/replications/rotenone_DA_neurons/MEMO.md) — Criterion D PASSED; A/B/C inverted |

---

## Tier 1 — Predictions 1 + 2 (this week, operator-personal compute)

These tests determine whether the hypothesis holds across actual ALS data + cardiomyocyte data. Both datasets are accessible without DUA.

| ID | Title | Status | Dataset | Time | Notes |
|---|---|---|---|---|---|
| **UR-1a** | **Test Prediction 1 on Kamath PD substantia nigra** — full-transcriptome PD DA neurons; resolves whether rotenone monotone-load was HVG-2000 artifact or real | **BLOCKED on FW-I2 personal GCP** | `/d/onedrive_mirror/.../kamath_pd/` (5 GB matrix.mtx.gz, 434k cells × 41k genes) | ~30-60 min on n2-highmem-32 | OOM on operator-personal Windows. **Critical test.** If Kamath shows peak-and-release after all, rotenone result was filter artifact and hypothesis revises. |
| **UR-1b** | **Test Prediction 1 on Pineda 2024 motor cortex** — actual ALS Betz cells (giant L5 pyramidal upper motor neurons) | OPEN | Synapse syn51105515 (~2-4 GB; needs free Synapse account) | ~30 min download + 1-2 hours compute | **THE definitive test.** ALS motor neurons specifically. If monotone-load reproduces here, hypothesis confirmed for the actual disease. |
| **UR-1c** | **Test Prediction 1 on Yadav 2023 spinal cord** — lower motor neurons in lumbar spinal cord | OPEN | GEO GSE190442 + GSE222322 | ~30 min download + 1 hour compute | Lower-MN complement to Pineda upper-MN. |
| **UR-2a** | **Test Prediction 2 on cardiomyocytes** | OPEN | `/d/onedrive_mirror/.../cellxgene/heart_atlas_59k.h5ad` + `heart_myocarditis_26k.h5ad` (already on D:) | ~1 hour | Both datasets cached locally; no download. |
| **UR-2b** | **Test Prediction 2 on heart failure cohort** | OPEN | GEO heart-failure scRNA series (e.g., GSE183852, GSE145154) | ~1 hour download + 1 hour compute | Disease-tissue cardiomyocytes vs healthy. |

**Critical path**: UR-1a (Kamath) is the gate. If Kamath PD reproduces the monotone-load pattern, Pineda + Yadav are confirmation. If Kamath shows peak-and-release, rotenone was the artifact and the hypothesis is wrong (or limited to specific stress modalities).

---

## Tier 2 — Prediction 3 (β-cell intermediate behavior)

| ID | Title | Status | Dataset | Time |
|---|---|---|---|---|
| **UR-3a** | **Test Prediction 3 on aged pancreas (β-cells)** | OPEN | `/d/onedrive_mirror/.../cellxgene/aging_pancreas.h5ad` (already on D:) | ~30 min compute | Tests whether β-cells (mixed proliferative-capable + post-mitotic) show mixed trajectory shape. |
| **UR-3b** | **Test Prediction 3 on type-2 diabetes pancreas cohorts** | OPEN | GEO T2D pancreas scRNA series | ~1 hour download + 1 hour compute | Disease-state β-cells. |
| **UR-3c** | **Stratify β-cell trajectory by proliferative-history markers** (Ki67/TOP2A/MCM2 history) | OPEN | Combined β-cell dataset post-UR-3a/3b | 1-2 hours | Direct test of Prediction 5 in β-cell context. |

---

## Tier 3 — Predictions 4-6 (longer arc; experimental design phase)

### Prediction 4: COPI manipulation in iPSC motor neurons

| ID | Title | Status | Time |
|---|---|---|---|
| **UR-4a** | Identify iPSC motor neuron differentiation protocol partner (academic collaboration target) | OPEN | 1-2 weeks outreach |
| **UR-4b** | Design COPI knockdown + overexpression experiment in iPSC-MN under stress | OPEN | 1-2 weeks experimental design |
| **UR-4c** | Run experiment + scRNA-seq trajectory + load-vs-release readout | DEFERRED | 6-12 months wet-lab |
| **UR-4d** | If validated, write therapeutic-implication paper | DEFERRED | post-UR-4c |

This is the causal test. Slow. Worth the time investment if Tier 1+2 predictions hold.

### Prediction 5: cycling-history correlation

| ID | Title | Status | Notes |
|---|---|---|---|
| **UR-5a** | Build cycling-history score (Ki67/TOP2A/MCM2 expression + per-cell aggregation) | OPEN | Methodology |
| **UR-5b** | Apply cycling-history-stratified trajectory analysis to mixed populations (neural crest derivatives, satellite cells + myocytes) | OPEN | After 5a |
| **UR-5c** | Check if cycling-history predicts trajectory shape | OPEN | After 5b |

### Prediction 6: tau-protective-against-load in cortical vs motor neurons

| ID | Title | Status | Notes |
|---|---|---|---|
| **UR-6a** | Quantify per-cell-type tau (MAPT) expression across cortical neurons + motor neurons + sensory neurons + spared neurons (oculomotor) | OPEN | Need full-transcriptome neuron datasets |
| **UR-6b** | Correlate tau levels with paraspeckle-load magnitude under matched stress | OPEN | After 6a |
| **UR-6c** | Test in tau KO vs tau OE iPSC neurons under stress | DEFERRED | Wet-lab |

---

## Tier 4 — Cross-tissue summary deliverable

| ID | Title | Status | Dependencies | Notes |
|---|---|---|---|---|
| **UR-X1** | **Cross-tissue trajectory-shape summary memo** | BLOCKED | UR-1a + UR-2a + UR-3a complete | The deliverable for Cai Lei + the methods-paper anchor: comparison of peak-vs-load across HUVEC (D03 done) + DA neurons (rotenone done) + PD substantia nigra (UR-1a) + ALS motor cortex (UR-1b) + spinal MN (UR-1c) + cardiomyocytes (UR-2a) + β-cells (UR-3a) |
| **UR-X2** | **Cai Lei first message attachment** (2-page) | BLOCKED on UR-X1 | The actual deliverable for Cai Lei outreach. Honest mechanism story; load-vs-release distinction; COPI leverage point. |
| **UR-X3** | **Patent provisional draft** for the load-vs-release diagnostic + COPI-restoration intervention method | BLOCKED on UR-X1 + counsel | Per [`37_Project_Sovereign/`](../37_Project_Sovereign/) IP architecture. |
| **UR-X4** | **Methods paper preprint** (LLC-authored, post-LLC formation) | BLOCKED on UR-X1 + LLC formation | Cross-tissue validation publication. |

---

## Critical-path summary

```
UR-1a (Kamath PD on personal GCP, ~$5-10 compute, ~1 day)
   ↓
   if monotone-load → hypothesis CONFIRMED on PD; proceed to UR-1b/1c/2a/3a in parallel
   if peak-and-release → hypothesis FALSIFIED; rotenone was HVG artifact; revert
   ↓
UR-1b + UR-1c + UR-2a + UR-3a (4 in parallel, ~1-2 days each)
   ↓
UR-X1 cross-tissue summary (1-2 days writeup)
   ↓
UR-X2 Cai Lei first message attachment (1 day)
   + UR-X3 patent provisional (counsel-led, weeks)
   + UR-X4 methods paper preprint (LLC-authored, weeks-months)
```

**Total Tier 1 + Tier 2 + UR-X1 + UR-X2 timeline**: ~2-3 weeks from FW-I2 personal GCP standup to Cai Lei-ready materials.

**Compute cost**: under $50 total on operator-personal GCP for the full Tier 1 + Tier 2 + UR-X1 batch.

---

## What this project does NOT do

- **Not the COPI manipulation experiment** (UR-4) — that's a separate project arc requiring wet-lab partnership, ~6-12 months
- **Not the tau-protection experiment** (UR-6) — also wet-lab dependent
- **Not Cai Lei outreach itself** — that lives in [37_Project_Sovereign/cai_lei/](../37_Project_Sovereign/cai_lei/); UR-X2 produces the attachment, the outreach decision is in Sovereign
- **Not the IP filing** — that's counsel-led + Sovereign LLC infrastructure; UR-X3 produces the technical content, the legal filing is elsewhere

---

## Scoring philosophy

Tier 1 (UR-1a especially) is the gate. If the load-vs-release hypothesis fails on Kamath PD, the rest of the project pauses pending hypothesis revision. **Honest negative is the expected outcome a meaningful fraction of the time** — that's what makes this an experimental program rather than a marketing exercise.

If Tier 1 confirms, Tiers 2+ run in parallel batches. Compute is cheap ($50 total batch); operator time is the bottleneck.

The endpoint deliverable is UR-X2 — a 2-page honest mechanism memo for Cai Lei. Everything else is supporting infrastructure for that single artifact.