# rare-disease-atlas.github.io — Rare Disease Venture & Translational Opportunity Atlas (v1.2)

> **2-Sentence Elevator Thesis:** **JAK2 V617F+ Myelofibrosis** is a $2.0B+ orphan blood cancer driven by a single point mutation in 50–60% of patients where zero approved drugs clear mutant stem cells or reverse bone marrow scar tissue (post-ruxolitinib median survival: 11–23 months). Pairing it with **Autoimmune Myelofibrosis (AIMF)**—a non-`JAK2` lupus-spectrum variant with dual transfusional/hepcidin iron buildup where marrow fibrosis **actually reverses in 74–80% of patients**—gives AI-bio, public-health, and venture builders both an asymmetric commercial wedge and *Nature's Positive Control* for curing human fibrosis.

## Canonical Problem-to-Startup Routing Table

| Code | Bottleneck / Clinical Failure Mode | Concrete Startup / Research Wedge | Target Track & Seed Budget |
| :--- | :--- | :--- | :--- |
| `[AI-BIO-01]` | All 4 approved JAKi drugs hit wild-type `JAK2` JH1, causing severe anemia/thrombocytopenia | Mutant-selective `JAK2 V617F` **JH2 allosteric degraders (PROTACs)** (`PDB: 4FVR`, `6E2Q`, `8BAK`) | AI / CompBio ($120K–$350K via CROs) |
| `[AI-BIO-02]` | Phase 3 combos (`MANIFEST-2`, `TRANSFORM-1`, `SENTRY`) hit `SVR35` but missed noisy `TSS50`; `IMpactMF` failed `OS` (Oct 2026) | **Composite FDA RDEA Surrogate Platform** (WSI reticulin pathomics + VAF + wearable ePROs) | AI / Public Health ($40K–$150K) |
| `[AI-BIO-03]` | Zero drugs reverse Grade 2/3 megakaryocyte-stromal marrow fibrosis | **scRNA-seq Perturbation Repurposing** (`GSE144568` + `GSE156644` `S100A8/A9`–`GLI1` loop) | AI / CompBio ($65K–$180K) |
| `[AI-BIO-04]` | Fibrosis assumed "irreversible" despite **74–80% histological resolution** in steroid/rituximab-treated **AIMF** | **Subtractive Spatial Multi-Omics** (resolving AIMF vs. persistent `JAK2+` PMF to drug `MMP`/`TIMP-1` switches) | AI / CompBio ($90K–$220K) |
| `[PUBLIC-HEALTH-01]` | >70% community-treated patients miss 2L trials due to late failure detection & "ruxolitinib washout terror" | **SMART-on-FHIR RR6 Failure Radar, External Control Arms & Zero-Washout Trial Matcher** | Public Health / RWE ($35K–$120K) |
| `[PUBLIC-HEALTH-02]` | Curative allo-HSCT reaches <10–15% of eligible patients due to community timing uncertainty | **AI Transplant Timing (`MIPSS70+` / `MTSS`) & Reimbursed RTM Cytopenia Concierge** | Public Health ($150K–$400K) |
| `[PUBLIC-HEALTH-03]` | 30–40% pathology discordance misdiagnoses curable `JAK2`-neg lupus/iron-overloaded **AIMF** as Triple-Negative PMF | **Multi-Modal Biopsy WSI + NGS + Lupus/Hepcidin-Iron Diagnostic Triage CDS** | Public Health / AI Path ($45K–$130K) |
| `[VENTURE-01]` | Academic & shelved MPN assets stall before $1.7B–$2.9B pharma M&A inflection | **Evergreen Venture-Philanthropy SPV Studio** stacking Blood Cancer United TAP + SBIR + $4.68M ODD fee waiver | Biotech Venture ($1.5M–$12M) |

## Primary Dossier & Funnel Spokes
- [00 // Main Autonomous Briefing Hub](https://rare-disease-atlas.github.io/): Overview, Skill-Routing Matrix, Paired `JAK2+` vs. `AIMF` Reversibility Table, Skeptical Venture Diligence Rebuttals, Clinical Trials, Open Datasets, and Existing Patient Communities.
- [01 // Rare Disease Business Models & Venture Economics](https://rare-disease-atlas.github.io/rare-disease-business-models.html): Cross-disease rubric, Orphan Drug Act unit economics, late-2020s Jakafi patent cliff, and SPV/venture-philanthropy models.
- [02 // Public Health, Clinical RWE & Care Navigation Wedges](https://rare-disease-atlas.github.io/public-health-biotech-wedges.html): Zero-wet-lab MPH/epidemiology ventures (`MPN PROGRESSion Registry®` `NCT07362225` ecosystem, External Control Arms, RR6 failure detection).
- [03 // AI & Computational Biology Drug Discovery Targets](https://rare-disease-atlas.github.io/ai-drug-repurposing-targets.html): JH2 V617F vs. WT JH2 selectivity, PROTAC exit vectors, and single-cell megakaryocyte–stromal fibrosis circuits (`GSE144568`, `GSE156644`, `GSE122198`, `GSE226340`).
- [04 // Autoimmune Myelofibrosis (AIMF), Diagnostic Triage & Iron/Hepcidin Axis](https://rare-disease-atlas.github.io/autoimmune-myelofibrosis-iron-wedge.html): Lupus-associated non-`JAK2` myelofibrosis, dual transfusional + IL-6/hepcidin iron trapping, preventing Triple-Negative PMF misdiagnosis, and *Nature's Positive Control* for reversing marrow fibrosis.

## Machine-Callable Endpoints, Open-Source Repos & 1-Command MCP Server
- **Run MCP Server Instantly (Zero PyPI Token Needed):**
  `uvx --from git+https://github.com/rare-disease-atlas/orphan-disease-opportunity-mcp orphan-disease-opportunity-mcp`
- [Standalone MCP Server Repo](https://github.com/rare-disease-atlas/orphan-disease-opportunity-mcp) · [Web Dossier Monorepo](https://github.com/rare-disease-atlas/rare-disease-atlas.github.io)
- [Full Consolidated Markdown Corpus (`/llms-full.txt`)](https://rare-disease-atlas.github.io/llms-full.txt): Complete single-fetch knowledge base including `JAK2+` PMF and Autoimmune Myelofibrosis + Iron/Hepcidin dossiers.
- [Structured Opportunity Atlas JSON (`/opportunities.json`, v1.2.0)](https://rare-disease-atlas.github.io/opportunities.json): 8 startup blueprints, 2 paired case studies (`jak2-myelofibrosis`, `autoimmune-myelofibrosis`), 7 grants, and 8 communities (`CC-BY-4.0`).
- **Persona-Filtered Edge JSON Slices (Low-Token `GET` Endpoints):** [AI / CompBio (`ai-bio.json`)](https://api.rare-disease-atlas.workers.dev/feeds/ai-bio.json) · [Public Health / RWE (`public-health.json`)](https://api.rare-disease-atlas.workers.dev/feeds/public-health.json) · [Biotech Venture (`venture.json`)](https://api.rare-disease-atlas.workers.dev/feeds/venture.json) · [Autoimmune MF & Iron (`aimf.json`)](https://api.rare-disease-atlas.workers.dev/feeds/aimf.json) · [Dynamic Filter (`/api/match?background=computational-ai-bio&wet_lab=false`)](https://api.rare-disease-atlas.workers.dev/api/match?background=computational-ai-bio&wet_lab=false)
- [Hugging Face Dataset (`rare-disease-atlas/rare-disease-venture-atlas`)](https://huggingface.co/datasets/rare-disease-atlas/rare-disease-venture-atlas) · [MLCommons Croissant 1.0 JSON-LD (`/croissant.json`)](https://rare-disease-atlas.github.io/croissant.json)
- [OpenAPI 3.1 Specification (`/openapi.yaml`)](https://rare-disease-atlas.github.io/openapi.yaml) · [MCP Autodiscovery (`/.well-known/mcp.json`)](https://rare-disease-atlas.github.io/.well-known/mcp.json) · [Official MCP Registry (`io.github.rare-disease-atlas/orphan-disease-opportunity-mcp`)](https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.rare-disease-atlas/orphan-disease-opportunity-mcp)

## Opt-In Agent Feedback, Missing Wedges & Citation Errata Protocol
- **Direct HTTP Webhook (`POST`, Zero Auth, Zero PII):** `POST https://api.rare-disease-atlas.workers.dev/api/feedback` with JSON `{"signal_type": "missing-wedge", "founder_background": "computational-ai-bio", "disease_id": "jak2-myelofibrosis", "agent_client": "...", "summary": "..."}`.
- **MCP Tool (`v1.3.0+`):** Call `submit_atlas_feedback(signal_type='missing-wedge' | 'citation-errata' | 'new-dataset' | 'disease-expansion', summary='...', founder_background='...', disease_id='...')` to report unserved founder backgrounds, updated `ClinicalTrials.gov` readouts, or newly deposited GEO/PDB datasets (strictly zero PII / zero confidential IP).
- **GitHub Issue / CLI Channel:** Coding agents with `gh` access or human builders can file structured `[agent-signal]` reports at <https://github.com/rare-disease-atlas/rare-disease-atlas.github.io/issues/new?template=agent_feedback.yml>.
- **Autonomous Agent Contributor Guide (`AGENTS.md` & `🤖🤖🤖` Fast-Track PRs):** <https://rare-disease-atlas.github.io/AGENTS.md> · [Standalone `SKILL.md` Repo](https://github.com/rare-disease-atlas/orphan-disease-venture-scout-skill)