Precision Medicine Mcp

Precision Medicine MCP Platform

[](https://www.python.org/downloads/) [](https://github.com/jlowin/fastmcp) [](https://modelcontextprotocol.io/) [](LICENSE)

> Dedicated to PatientOne -- a dear friend who passed from High-Grade Serous Ovarian Carcinoma in 2025.

This platform automates multi-modal data processing for clinical decision support — all results require clinician review before any clinical action.

---

The Problem

Standard oncology workup (BRCA1/2, HRD panel, tumor genomic panel) generates no immunotherapy or investigational hypotheses. For preventive health, standard lipid panels and population genetic screens miss key risk factors. Manual multi-modal analysis across genomics, spatial transcriptomics, and clinical data is clinically impractical -- the platform automates it.

The Platform

A multi-server MCP architecture orchestrated by AI (Claude + Gemini) executes a 5-stage pipeline:

flowchart LR
    Z["0 De-IDPreprocessing"] --> A["1 DataAcquisition"]
    A --> B["2 SpatialDeconvolution"]
    B --> C["3 TargetProfiling"]
    C --> D["4 CausalInference"]
    D --> E["5 Report"]

    subgraph servers [" "]
        direction TB
        S0["De-identify · JSON · PDF · DOCX · VCF"]
        S1["EHR · GEO · TCGA"]
        S2["Spatial · DeepCell · CIBERSORTx"]
        S3["OpenTargets · Neoantigen"]
        S4["Perturbation · Quantum"]
        S5["Patient Report"]
    end

    Z --- S0
    A --- S1
    B --- S2
    C --- S3
    D --- S4
    E --- S5

    AI(["AI OrchestratorClaude + Gemini"]) -.-> Z
    AI -.-> A
    AI -.-> B
    AI -.-> C
    AI -.-> D
    AI -.-> E

Architecture at a glance

                  +--------------------------------------+
                  |           CLIENT LAYER               |
                  |  Claude Desktop / Hospital EHR       |
                  |  Adapter / Research Notebook         |
                  +----------------+-----------------+
                                   |
                         MCP (FastMCP >= 2.13)
                                   |
   +---------------------------------------------------------------+
   |                                                               |
   |  PRE-PROCESSING (Stage 0)                                     |
   |  deidentify                                                   |
   |                                                               |
   |  DATA ACQUISITION      ANALYSIS & INFERENCE      REPORTING   |
   |                                                               |
   |  mockepic              spatialtools              patient-     |
   |  epic                  multiomics                report       |
   |  geodownload           perturbation                           |
   |  mocktcga              quantum-fidelity                       |
   |  genomic-results       opentargets                            |
   |  fgbio                 neoantigen                             |
   |                        cibersortx                             |
   |                        openimagedata                          |
   |                        deepcell                               |
   |                        cell-classify                          |
   |                        cardiometabolic                        |
   +---------------------------------------------------------------+

All tools accessible via natural language. Every AI result requires clinician APPROVE/REVISE/REJECT. HIPAA-compliant. Current server and tool counts: [Server Registry](docs/reference/shared/server-registry.md).

The Results

The platform surfaces clinically actionable findings that standard workup cannot reach — 6 investigational hypotheses across 2 cancer types plus 3 preventive health evidence gaps, validated across three independent use cases:

| Use Case | Patient | Key Finding Missed by Standard Workup | |---|---|---| | HGSOC (Stage IV) | PAT001 | 3 investigational paths: neoantigen vaccine (RMPEAAPPV IC50 7.8 nM), NNMT/CAF inhibition, convergent checkpoint blockade | | ER+ Breast Cancer | PAT002 | 3 investigational hypotheses: inavolisib over alpelisib (PIK3CA H1047R, 2024 FDA approval), MYC-driven triple therapy, YSAPLSSSL neoepitope vaccine + CAF depletion + anti-PD-1 — zero disease-specific code changes | | Preventive Cardiovascular | PAT003 | Intermediate CVD risk (Reynolds 14.3%) with 3 high-priority gaps missed by standard lipid panel AND population genetic screen: Lp(a), APOE genotype, CAC score |

The same 20-server architecture runs all three. No disease-specific code changes between use cases.

Validated results — PAT001 (HGSOC)

| Metric | Value | Source server | |--------|-------|---------------| | HRD score | 54 | mcp-genomic-results | | TMB (POLE-corrected) | 47.3 mut/Mb | mcp-genomic-results | | Top neoantigen IC50 (RMPEAAPPV) | 7.8 nM | mcp-neoantigen | | Spatial spot count | 900 | mcp-spatialtools | | Moran's I (global) | -0.0033 | mcp-spatialtools | | Deconvolution: tumor | 56 cells | mcp-cibersortx | | Deconvolution: endothelial | 44 cells | mcp-cibersortx | | Deconvolution: macrophages | 43 cells | mcp-cibersortx | | Deconvolution: fibroblasts | 41 cells | mcp-cibersortx | | Deconvolution: CD8+ T cells | 30 cells | mcp-cibersortx |

---

Try It

# Clone and explore
git clone https://github.com/lynnlangit/precision-medicine-mcp.git
cd precision-medicine-mcp

# Run tests for any server (DRY_RUN mode, no external deps needed)
cd servers/mcp-multiomics && uv run pytest -v

# Or use Claude Code to explore interactively
claude

All servers default to DRY_RUN mode (mock responses, no API keys needed) for quick validation. Set *_DRY_RUN=false to use synthetic patient data for end-to-end testing.

---

Learn More

| Audience | Start Here | |----------|------------| | Getting Started | [Installation Guide](docs/getting-started/installation.md) | | Funders | [Executive Summary](docs/for-funders/EXECUTIVE_SUMMARY.md) | | Hospitals | [Hospital Guide](docs/for-hospitals/README.md) | | Developers | [Architecture](docs/for-developers/ARCHITECTURE.md) | | Researchers | [Researcher Guide](docs/for-researchers/README.md) | | Educators | [Educator Guide](docs/for-educators/README.md) | | All docs | [Documentation Index](docs/INDEX.md) |

Video: 5-minute demo | Paper: [Why MCP for Healthcare](docs/reference/architecture/WHYMCPFORHEALTHCARE.md) | External connectors: [Setup guide](docs/for-researchers/CONNECTEXTERNALMCP.md)

---

Apache 2.0 | Python 3.11+ | FastMCP >= 2.13 | uv for package management

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: lynnlangit
• Source: lynnlangit/precision-medicine-mcp
• License: Apache-2.0

Install and usage instructions live in the source repository linked above.