NCLEdetector 2.0.0

Non-covalent Lasso-like Entanglement (NCLE) Detection in Protein Structures and trajectories

NCLEdetector

A comprehensive Python package for studying non-covalent lasso entanglements in protein folding through molecular dynamics simulations and experimental data analysis.

Overview

NCLEdetector provides a complete toolkit for analyzing protein entanglements across multiple scales - from individual structures to large-scale proteomic datasets. The package enables researchers to:

• Identify and characterize native entanglements in protein structures
• Calculate order parameters for simulation trajectories (Q, G, K, SASA, Jwalk, XP)
• Build Markov State Models from coarse-grained simulation ensembles
• Compare simulations to experiments using LiP-MS and XL-MS data
• Perform population-level analysis across heterogeneous protein datasets
• Coarse-grain and back-map between atomic resolutions

Key Features

• Multi-scale Analysis: From single proteins to proteome-wide studies
• Experimental Integration: Direct comparison with mass spectrometry data
• Advanced Clustering: Non-native entanglement clustering and MSM construction
• Statistical Methods: Monte Carlo simulations and logistic regression modeling
• Flexible Resolution: Seamless conversion between all-atom and coarse-grained representations

Installation

NCLEdetector supports two installation methods. Choose the one that best matches your workflow.

Method 1: Conda environment install (Linux and macOS)

Use this when you want a local Python environment and direct script/package development.

From the repo root:

conda env create -f environment.yml
conda activate ncledetector

Notes:

• The provided conda environment targets Python 3.11 for best compatibility with the scientific stack.
• Run conda env create from the NCLEdetector repo root (the environment file uses pip -e .).
• If you prefer installing into an existing env, use pip install -e . from the repo root.

macOS notes:

• The default environment.yml is intended to work on both Linux and macOS via conda-forge.
• If you hit solver/build issues on macOS, try:

conda env create -f environment-mac.yml
conda activate ncledetector

• On Apple Silicon, if a dependency is missing for osx-arm64, a common workaround is:

CONDA_SUBDIR=osx-64 conda env create -f environment-mac.yml
conda activate ncledetector

Method 2: Container install (Docker and Apptainer/Singularity)

Use this for reproducible runs, HPC workflows, or when you do not want to manage local dependencies.

Pull prebuilt images:

# Docker
docker pull ghcr.io/obrien-lab-psu/ncledetector:latest

# Apptainer/Singularity
apptainer pull ncledetector.sif docker://ghcr.io/obrien-lab-psu/ncledetector:latest

For full run examples and bind-mount guidance, see Documentation/container_usage.md.

Tutorials

Step-by-step, runnable tutorials covering all four analysis workflows are in Documentation/. Start here:

• Documentation/index.md — master index with environment setup, path variables, and links to all workflows

For quick CLI reference, every script supports --help:

python scripts/run_nativeNCLE.py --help
python scripts/run_OP_on_simulation_traj.py --help

Package Structure

NCLEdetector/
├── NCLEdetector/                    # Main package
│   ├── __init__.py
│   ├── gaussian_entanglement.py  # Core entanglement calculations
│   ├── clustering.py             # Entanglement clustering methods
│   ├── order_params.py          # Order parameter calculations
│   ├── compare_sim2exp.py       # Simulation-experiment comparison
│   ├── statistics.py            # Statistical analysis methods
│   ├── entanglement_features.py # Feature generation
│   ├── change_resolution.py     # Resolution conversion
│   └── utilities.py             # Helper functions
├── scripts/                     # Example workflow scripts
├── Documentation/               # Detailed module documentation
└── TestingGrounds/             # Test data and examples

Core Modules

• gaussian_entanglement: Calculate Gaussian linking numbers and identify entanglements
• clustering: Cluster native and non-native entanglements, build MSMs
• order_params: Compute Q, G, K, SASA, Jwalk, and cross-linking propensity
• compare_sim2exp: Integrate LiP-MS and XL-MS experimental data
• statistics: Population modeling and Monte Carlo analysis
• entanglement_features: Generate structural features for entanglements
• change_resolution: Convert between all-atom and coarse-grained representations

Documentation

Detailed documentation for each module:

• Tutorial index
• Workflow 1: Native NCLE detection
• Workflow 2: Trajectory analysis
• Workflow 3: Sim-to-experiment comparison
• Workflow 4: Population-level analysis
• Container usage (Docker and Apptainer/Singularity)
• Gaussian Entanglement
• Clustering
• Order Parameters
• Simulation-Experiment Comparison
• Statistical Analysis
• Entanglement Features
• Resolution Conversion
• Utilities

Requirements

• Python 3.8+
• NumPy
• Pandas
• SciPy
• MDAnalysis
• OpenMM (for force field operations)
• Matplotlib (for visualization)
• See environment.yml for complete dependencies

Citation

If you use NCLEdetector in your research, please cite:

@software{ncledetector2024,
  title={NCLEdetector: A Python Package for Protein Entanglement Analysis},
  author={Ian Sitarik},
  author={Yang Jiang},
  author={Hyebin Song},
  author={Edward O'Brien},
  year={2026},
  url={https://github.com/obrien-lab-psu/NCLEdetector}
}

Contributing

Contributions are welcome! Please see our contributing guidelines and submit pull requests for any improvements.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

For questions and support, please open an issue on GitHub or contact the developers.