MyNHANES is an open-source data platform designed to simplify access, integration, and transformation of public health data from the National Health and Nutrition Examination Survey (NHANES).
Unlike the official NHANES portal — where datasets are fragmented across cycles and require manual download and parsing — MyNHANES offers a structured environment to ingest, transform, standardize, and query NHANES data as a unified relational database.
- ✅ Unified access to raw data across NHANES cycles
- 🔍 Full metadata integration and search via variables, tags, and cycles
- 🧪 Rule-based transformation engine for derived variables
- 🧭 Web-based interface for non-programmers (Django Admin)
- 📦 Export-ready data for analysis (CSV)
- 📖 Transparent logs and versioning of all steps
- 🔄 Syncs with public GitHub repository for master data and rule definitions
pip install mynhanesAfter installation, use the CLI with enviroment activated:
mynhanes deploy --type local
mynhanes runserverThen access the admin interface at: http://127.0.0.1:8000/admin
# Deploy local DB and import master data
mynhanes deploy --type local
# Run ingestion for marked datasets
mynhanes ingestion_nhanes --type db
# Apply active transformation rules
mynhanes transformation
# Export query reports
mynhanes query --fields_report- Imported from
MyNHANES_DataHub - Includes: Cycles, Datasets, Variables, Tags, Rules
- Downloads
.XPTand.HTMfiles - Parses metadata and values
- Stores data in the local relational model (
Data)
- Each rule is a Python script mapped to input/output variables
- Rules run automatically or manually via admin/CLI
- Results are versioned and saved alongside raw data
- Use QueryStructures and QueryColumns to define reusable queries
- Export structured CSVs across cycles with selected variables
You want to identify NHANES participants likely to have Parkinson’s Disease based on their medication.
With MyNHANES, you can:
- Enable relevant datasets (
RXQ_RX,DEMO) - Apply a transformation rule (
rule_parkinson) - Export the final dataset for modeling
All steps are documented, logged, and reproducible: https://github.com/HallLab/MyNHANES/tree/main/rules/rule_00001
MyNHANES is built to support research that demands:
- Harmonized datasets across cycles
- Rapid exploration and export of clean variables
- Transparent data derivation pipelines
- Integration with tools like R, Python, or Jupyter Notebooks
It empowers data scientists, epidemiologists, and clinical researchers to focus on analysis, not manual preprocessing.
We welcome contributions from the community!
- 📥 Submit issues or feature requests
- 🧪 Contribute transformation rules to the
rules/folder - 🗂 Help improve tagging and variable descriptions
- 📊 Share notebooks or case studies using MyNHANES data
- Core App: https://github.com/HallLab/MyNHANES
- DataHub: https://github.com/HallLab/MyNHANES_DataHub
This project is licensed under the MIT License.
For questions or collaboration inquiries, please contact: Andre Rico – [email protected] or open an issue in the GitHub repository.