bngsim 0.11.33

Embeddable simulation engine for BioNetGen reaction networks

bngsim

Embeddable simulation engine for BioNetGen reaction networks.

bngsim is a high-performance C++ simulation kernel with Python bindings that replaces BioNetGen's subprocess-based run_network driver. It loads .net files, runs ODE (CVODE) and SSA (Gillespie) simulations in-process, and returns results as NumPy arrays — no file I/O, no subprocess spawning, no Perl dependency.

Highlights

• Fast — in-process execution with the GIL released during simulation; thread-parallel batch sweeps
• No toolchain at runtime — import bngsim; no compilers, no BNGPATH, no Perl
• Modern SUNDIALS — v7.x CVODE/CVODES with re-entrant SUNContext
• Multi-format — loads .net, Antimony (.ant), and SBML (.xml) models
• Rich results — NumPy arrays, named observable access, pandas/xarray, HDF5 save/load
• Standards interchange — .net/cBNGL ⇄ SBML and SED-ML + OMEX packaging, every conversion verified faithful
• Gradient-ready — CVODES forward sensitivities, Fisher information, JAX-differentiable RHS
• Validated — matches run_network to ~10⁻¹² (ODE) and cross-checked against RoadRunner and the SBML semantic test suite

Installation

pip install bngsim

Prebuilt wheels ship for common platforms; large / genome-scale models benefit from the optional sparse (KLU) solver. See the installation guide for source builds, the KLU dependency, and optional extras.

Quickstart

import bngsim

# Load a model (.net, Antimony, or SBML) and run an ODE simulation
model = bngsim.Model.from_net_file("model.net")
sim = bngsim.Simulator(model, method="ode")
result = sim.run(t_end=100.0, n_steps=1000)

result.times          # (1001,) NumPy array of time points
result["A"]           # trajectory of observable "A"
result.to_dataframe() # pandas DataFrame of all observables

Stochastic (method="ssa" / "psa"), network-free (NFsim / RuleMonkey), sensitivity analysis, steady-state, and events are all covered in the quickstart and user guide.

Documentation

Full documentation is hosted at bngsim.readthedocs.io and lives in docs/:

• Installation · Quickstart
• User guide — loading models, simulation, network-free, results, events, sensitivities, steady state, SBML/SED-ML/OMEX interchange
• Reference — API · expression language
• Architecture · benchmarks & validation

Benchmarks & validation

bngsim beats run_network on every SSA/PSA model measured (geometric-mean speedup ~8.7× SSA), agrees with libRoadRunner across the full BioModels ODE corpus, and scores 1577 Match on the SBML semantic test suite with zero wrong-but-plausible answers. The numbers, methodology, and reproducible harnesses are in benchmarks & validation, with the committed cross-engine snapshots under parity_checks/. To re-run the suites and reproduce the numbers — the pinned-fetch model, corpus provenance, tool versions, and a no-download smoke path — see REPRODUCING.md. The supported-construct matrix is in SUPPORT_MATRIX.md.

Contributing

Build-from-source, test, and CI instructions are in CONTRIBUTING.md and the development docs. Release history is in CHANGELOG.md.

License

MIT. See LICENSE for the full text and the Triad/LANL copyright notice.

Third-party code and model/test-data corpora that BNGsim redistributes are listed, with their licenses, in NOTICE.

© 2026. Triad National Security, LLC. All rights reserved. This is a Los Alamos National Laboratory open-source release; LANL software release reference O5098.

Citation

If you use bngsim in your research, please cite:

Preprint coming soon!

Acknowledgments

BNGsim was developed and validated against RoadRunner, AMICI, COPASI, BioNetGen, and the SBML/DSMTS test suites. See ACKNOWLEDGMENTS.md for citations.