tangle

       ╔══════════╗
  ┌────╢  ●──●──●  ╟────┐
  │    ║   ╲ │ ╱   ║    │
  │    ║  ●─●●─●   ║    │
  └────╢   ╱ │ ╲   ╟────┘
       ╚══════════╝

tangle is a high-performance C++ library for static protein–protein interaction (PPI) network analysis, designed to support drug discovery and systems biology. It offers a clean C++ API, a powerful CLI, and an interactive TUI.

Installation

macOS (Homebrew)

brew tap sulkysubject37/tangle
brew install sulkysubject37/tangle/tangle

This installs the tangle CLI, the tangle-tui TUI, and the C++ library/headers into your Homebrew prefix.

From source (Linux & macOS)

mkdir build && cd build
cmake ..
make

You can then run the binaries from the build directory (./tangle, ./tangle-tui) or install system-wide via cmake --install . / sudo make install.

Generic Linux tarball

On a Linux machine, you can build a self-contained tarball (works across most distros):

mkdir build && cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
cmake --build .
cpack -G TGZ

This produces a file like tangle-<version>-Linux.tar.gz containing bin/, lib/, and include/. Users can unpack it anywhere and add the bin/ directory to their PATH.

Quirk: tangle is written in C++ because large, static PPI networks tend to make Python and R sweat — they pay extra overhead for dynamic typing, copying data frames, and interpreter/GIL costs, while tangle can stream tens of millions of edges with tight memory and cache-friendly layouts.

1. Core Library

Built for speed and memory efficiency, the tangle library provides:

• Graph Engine: Optimized adjacency lists for large scale networks (STRING, BioGRID).
• Algorithms:
  • Centrality: Degree centrality.
  • Communities: Louvain community detection.
• Enrichment: Hypergeometric GO enrichment analysis.
  • Optimized: 1000x faster than standard implementations via pre-computed frequency maps.
  • Smart IDs: Supports both UniProt IDs and Gene Symbols (e.g., "FGF1" matches "P05230").
• I/O: robust importers for PPI standards and SBML export.

2. CLI (tangle)

The command-line interface exposes the library functions for scripting and pipelines.

Usage

# Import a STRING network
tangle import --in=9606.protein.links.txt --out=human.edgelist --score=700

# Run analysis (Louvain clustering)
tangle analyze --in=human.edgelist --out=communities.tsv

# Run GO Enrichment
tangle annotate --in-comm=communities.tsv --in-gaf=goa_human.gaf --out=enrichment.tsv

# Export to SBML
tangle export --in=human.edgelist --out=network.sbml

3. TUI (tangle-tui)

An interactive Text User Interface (using FTXUI) for exploring networks without writing code.

Key Features

• Tabbed Interface: Organized workflow: Import -> Analysis -> Annotation -> Export.
• One-Click Enrichment:
  • Automatically loads GAF annotation files when needed.
  • Instant analysis of communities.
  • Save Results: Export your findings to TSV with a single click.
• Robust & Thread-Safe:
  • Long-running tasks happen in the background.
  • Loading spinners and real-time logs keep you informed.
• Interactive:
  • Preview input files before loading.
  • User-friendly pagination prevents UI lag.

4. End-to-End Pipeline Guide

Follow this guide to run a complete analysis pipeline using the provided test data.

1. Import Raw Data

Convert the raw STRING-like data (test_data/test_case.tsv) into Tangle's optimized edgelist format.

./tangle import --in=test_data/test_case.tsv --out=graph.edgelist --format=string --score=400 --score-col=12 --delimiter=tab

2. Community Detection

Identify functional modules (communities) using the Louvain algorithm.

./tangle analyze --in=graph.edgelist --out=communities.tsv

3. Functional Enrichment

Annotate the discovered communities using Gene Ontology (GO) terms.

# Using the provided dummy GAF for testing (or provide your own GAF file)
./tangle annotate --in-comm=communities.tsv --in-gaf=tests/dummy_goa.gaf --out=enrichment_results.tsv

4. Export to SBML

Export the network structure to SBML for use in other systems biology tools.

./tangle export --in=graph.edgelist --out=network.sbml

Getting Started

Build from source

If you did not install via Homebrew, you can build from source as follows:

mkdir build && cd build
cmake ..
make

Run

From a source build, run the TUI with:

./tangle-tui

If installed via Homebrew, you can launch it directly with:

tangle-tui

Citation

MD. Arshad, Department of Computer Science, Jamia Millia Islamia https://doi.org/10.5281/zenodo.17927401

Blog

Link : https://subconc.hashnode.dev/tangle