RosettaCommons/rosetta

The Rosetta Bio-macromolecule modeling package. Available through license with the University of Washington.

What it solves

Rosetta provides a comprehensive suite of algorithms for the computational modeling and analysis of protein structures. It addresses complex challenges in computational biology, such as predicting the structure of biological macromolecules and designing new proteins or enzymes from scratch (de novo design).

How it works

Rosetta uses a large codebase of C++ algorithms to simulate and analyze the physical and chemical properties of proteins. It can be used as a standalone software suite, compiled from source, or accessed via PyRosetta, which provides Python bindings to the library for easier scripting and integration. The system also supports integration with deep learning frameworks like libtorch and TensorFlow through specific Docker builds.

Who it’s for

It is designed for researchers in computational biology and biochemistry who need to perform ligand docking, protein design, or macromolecular complex structure prediction.

Highlights

  • Broad Application: Supports de novo protein design, enzyme design, and ligand docking.
  • PyRosetta: Offers Python bindings to make the powerful C++ library accessible to Python developers.
  • Flexible Deployment: Available via official Docker images or Conda packages for rapid setup.
  • Extensive Ecosystem: Maintained by a global collaboration of over 100 academic research groups.

Related

  • Project
    biotite-dev/biotite

    A Python library for computational molecular biology that provides tools for analyzing and manipulating biological sequence and structure data using NumPy for high performance.

  • Project
    RosettaCommons/foundry

    Foundry is a tooling and infrastructure framework for protein design, providing a unified interface to train and use models for generative design, structure prediction, and inverse folding.

  • Project
    martinpacesa/BindCraft

    An automated pipeline for de novo binder design that uses AlphaFold2 backpropagation, ProteinMPNN, and PyRosetta to create proteins that bind to specific targets.

  • Project
    bytedance/Protenix

    Protenix is an open‑source deep‑learning system for high‑accuracy protein (and RNA‑protein) structure prediction. It builds on AlphaFold 3, offers several model sizes (including a fast Mini version), supports MSA/template/RNA inputs, and adds constraint‑aware folding. The package is installable via pip, includes a CLI, and a public web server. Benchmarks claim it outperforms AlphaFold 3 on antibody‑antigen complexes. All code is Apache 2.0; only the newest v2 weights are proprietary.

  • Project
    aurekaresearch/OpenDDE

    OpenDDE is an open‑source, all‑atom biomolecular foundation model that predicts structures of proteins, DNA, RNA, ligands, and ions by co‑folding them together. It offers conditional (constraint‑guided) inference, multi‑GPU Fold‑CP scaling, and optional MSA/template preprocessing. Install via PyPI, source, or Docker; run predictions with a JSON job file using the `opendde pred` CLI. The project is in preview, with checkpoints hosted on Hugging Face and an Apache‑2.0 license.