STARMAP is a computational framework for integrating protein structure, mutation data, transcriptional networks, and drug response to identify functionally relevant regions of proteins and their downstream effects.
This repository contains code associated with the STARMAP manuscript (currently in preparation), including reproducible analysis workflows and example pipelines.
This project is released under a custom restricted use license.
- Free for non-commercial academic and research use
- Modification and redistribution allowed for academic purposes
- Commercial use is strictly prohibited without explicit permission
For commercial licensing or other inquiries, please contact elizabeth_brunk@med.unc.edu.
A minimal working pipeline is provided for TP53 to demonstrate how STARMAP can be run end-to-end. This includes:
- Structural processing and feature generation
- Dimensionality reduction and clustering (Regions of Functional Interest, RFIs)
- Transcriptional regulatory network (TRN) association analysis
- Drug response modeling
This serves as a reference implementation for applying STARMAP to additional proteins.
Code used to generate the analyses and figures from the manuscript is included. These scripts:
- Reproduce key statistical analyses
- Generate plots and summary metrics
- Recreate figure panels from the paper
Note that some analyses rely on large intermediate datasets and may require adaptation depending on compute environment.
Due to the size of the full STARMAP outputs, they are not hosted directly in this repository.
Full processed outputs, including:
- Protein-level scores
- Cluster-level annotations
- TRN associations
- Drug response metrics
are available at:
The pipeline relies on standard scientific Python libraries, including:
- pandas
- numpy
- scikit-learn
- matplotlib / seaborn
- scipy
Additional dependencies may be required for specific modules.
To run the TP53 pipeline:
- Navigate to the TP53 example directory
- Follow the provided script/notebook instructions
- Ensure required input data paths are configured
The STARMAP manuscript is currently in preparation.
A citation will be provided upon publication.
For questions, collaboration inquiries, or licensing requests, please reach out to the repository authors.