fuzzyfold

A high-performance framework for stochastic RNA folding kinetics.

fuzzyfold is an open-source Rust workspace for nucleic acid secondary structure analysis with an explicit focus on kinetic modeling. It combines a Gillespie-type stochastic simulation engine with a flexible move set (including three-way and four-way shift moves) and provides both command-line tools for end users and a library interface with Python bindings for custom workflows.

Features

Stochastic simulation of RNA and DNA secondary structure folding at base-pair resolution
Three-way and four-way shift moves with Arrhenius-based activation energy parameterization
Ensemble-level occupancy analysis via aggregation of parallel trajectories
Co-transcriptional folding trajectories
Thermodynamic parameters: RNA (Turner 2004), DNA (RNAstructure), extended sets including modification parameters and special hairpin parameters
Energy evaluation fully consistent with ViennaRNA
Pure Rust workspace: straightforward installation via cargo, reproducible builds, no external runtime dependencies
Python bindings for rapid prototyping and custom analysis workflows

Installation

fuzzyfold requires a working Rust toolchain. To install all command-line tools:

cargo install fuzzyfold

Individual crates can also be installed separately:

cargo install ff_structure
cargo install ff_energy
cargo install ff_kinetics

Command-line tools

Program	Description
`ff-trajectory`	Single stochastic folding trajectory
`ff-timecourse`	Ensemble occupancy analysis over multiple parallel trajectories
`ff-explore`	Enumerate secondary structure neighborhoods and macrostates
`ff-eval`	Free-energy evaluation for secondary structures
`ff-randseq`	Generate a random nucleic acid sequence

See directory examples for basic usage of command-line software, as well as examples/workflows to reproduce published analyses.

Crates

Crate	Description
`ff_structure`	Nucleic acid secondary structure data structures
`ff_energy`	Free-energy evaluation using nearest-neighbor parameters
`ff_kinetics`	Stochastic folding kinetics engine
`fuzzyfold`	Command-line interfaces

Additional crates are in development and will be published to crates.io as they mature.

Python bindings

Python bindings are available for all major simulation and analysis components, enabling rapid prototyping before contributing high-performance Rust code. See the crates/ff_python directory for examples and installation instructions.

Repository structure

Branch	Description
`main`	Stable, well-documented, publication-ready code
`development`	Integrated development branch; some experimental features
`dev_feature`	Early crate or feature proposals

Contributing

Contributions are welcome! Please start with small contributions, such as bug fixes, performance improvements, documentation, or examples. For new features or new crates associated with fuzzyfold, please reach out before submitting a pull request. The goal of the workspace is a coherent, well-documented ecosystem for RNA and DNA secondary-structure modeling and kinetic simulation, with each crate covering a clearly separated aspect of the workflow.

To run the benchmark suite:

cargo bench --workspace

Please open a GitHub issue to discuss larger changes before submitting a pull request.

Citation

If you use fuzzyfold in your research, please cite:

Stefan Badelt (2026): fuzzyfold: a high-performance framework for stochastic RNA folding kinetics, [https://doi.org/]

License

MIT — see LICENSE for details.

ff_structure 0.3.1