PyRosetta is moving to quarterly builds!
Downloads on this page are provided to academic and non-commercial users under the PyRosetta Software Non-Commercial License Agreement. This license applies to all downloads on this page.
Use for commercial purposes requires purchase of a separate license. This includes fee-for-service work by academic users. More information is available here through the University of Washington's CoMotion office or you can email license@uw.edu for more information.
PyRosetta is available for Intel-based Mac's (64-bit OS X v10.5+) and M1-arm64 Mac's as well as on Linux (64-bit Red Hat and 64-bit Ubuntu) platforms in continuous release versions (usually we have a fresh builds available daily). Running on Windows-10 (thought Unix layer) is are also supported. In order to run PyRosetta you will need to use Python-3.6 or later. If you would like to build PyRosetta from source please see relevant section in the Documentation page.
For technical support and assistance, please open a Discussion or Issue in the associated GitHub repository or post on Slack (Rosetta Commons members only.) Our forums are no longer accepting new posts, but have a wealth of information and we recommend you search for an answer to your question there as well.
PyRosetta binaries is available in three different forms: (a) as Python packages (recommended), (b) Python-wheel packages and (c) as Conda packages. We also provide auxiliary pyrosetta-installer pip package to help install the PyRosetta weekly builds. Installing PyRosetta as a Python package (recommended) will install the stable quarterly build that will be hosted long-term to allow for the reproducibility of results.
Persistent wheel packages will now be maintained long-term, allowing for the straightforward creation of reproducible PyRosetta environments across all supported builds and Python versions.
Release - Speed optimized build, use it for production runs.
MinSizeRel - Build optimized to reduce memory footprint. Use this build with low-memory systems.
Debug - Binaries compiled in debug mode with additional asserts enabled and with debug-info compiled-in. Use this build for debugging.
Python-x.x versions - PyRosetta build for specific Python version
The following commands install the long-term supported, quarterly PyRosetta builds. There are two mirrors to choose from, West or East, which contain the same packages but the download speed may vary between the two of them depending on your location.
Quarterly US West coast mirror (if unsure, use this mirror):
Standard release (maximally optimized for performance):
pip install pyrosetta --find-links https://west.rosettacommons.org/pyrosetta/quarterly/release
MinSizeRel (smallest possible binary size, minor performance drawbacks):
pip install pyrosetta --find-links https://west.rosettacommons.org/pyrosetta/quarterly/minsizerel
Thread serialized:
pip install pyrosetta --find-links https://west.rosettacommons.org/pyrosetta/quarterly/release.cxx11thread.serialization
Quarterly US East coast mirror:
Standard release (maximally optimized for performance):
pip install pyrosetta --find-links https://graylab.jhu.edu/download/PyRosetta4/archive/release-quarterly/release
MinSizeRel (smallest possible binary size, minor performance drawbacks):
pip install pyrosetta --find-links https://graylab.jhu.edu/download/PyRosetta4/archive/release-quarterly/minsizerel
Thread serialized:
pip install pyrosetta --find-links https://graylab.jhu.edu/download/PyRosetta4/archive/release-quarterly/release.cxx11thread.serialization/
The pyrosetta-installer package was originally created due to the large size of the PyRosetta package. The installation method described below will still work, however, it will install the 'devel' (weekly) builds. These builds are not kept indefinitely, so they cannot be used to create long-term, reproducible results.
pip install pyrosetta-installer
python -c 'import pyrosetta_installer; pyrosetta_installer.install_pyrosetta()'
The install_pyrosetta function also takes few optional arguments:
distributed=False - install PyRosetta with cxx11thread.serialization support and also install packages required for pyrosetta.distributed framework
extras='' - directly supply dot separated list of of Rosetta C++ extras flags
mirror=0 - mirror to use, default is to use west mirror
serialization=False - install PyRosetta cxx11thread.serialization build
silent=False - minimize log output during installation
skip_if_installed=True - skip install if PyRosetta install detected
type='Release' - PyRosetta build to install: Release, MinSizeRel
use_setup_py_package=False - use PyRosetta setup.py package instead of wheel. Could be useful in cases when wheel install fails. Note that by default installer will fallback to use this option if an attempt to install a wheel fails.
For up-to-date list of supported options see https://pypi.org/project/pyrosetta-installer
pip install pyrosetta-distributed
For up-to-date list of supported options please see https://pypi.org/project/pyrosetta-distributed
[Devel US East coast mirror] [Devel US West coast mirror]
PyRosetta-4 Linux build work natively thought Win10 Linux layer. For installation details please see: Instructions for Installing PyRosetta on Windows 10
[PyMOL-RosettaServer.python2] [PyMOL-RosettaServer.python3]
[PyMOL-Rosetta-relay-client.python2][PyMOL-Rosetta-relay-client.python3]
The RosettaCommons conda channel provides conda packages for each weekly release. Currently PyRosetta.release packages for Python 3.6-3.14 are provided for both Mac and Linux platforms. Note that when possible we intend to keep all released packages available (i.e old releases packages will not be deleted), so it should be safe to publish your results and use explicit PyRosetta version to ensure reproducibility.
To download PyRosetta via conda use one of the following commands:
For those on the WEST cost (if unsure, use this mirror):
conda install -y -c https://conda.rosettacommons.org -c conda-forge pyrosetta
For those on the EAST cost:
conda install -c https://conda.graylab.jhu.edu -c conda-forge pyrosetta
You should be able to use either mirror (conda.rosettacommons.org or conda.graylab.jhu.edu) you will likely just find that one is faster than the other based on your location.
Our release archive: [full list of available releases]
PyRosetta is distributed as a standard Python "distutils" package, which is compatible with most Python package management systems. (Though due to licensing issues, it is not available through pip.) There are two major ways to install PyRosetta: either into your standard system Python, or using a Python environment manager. Installation into the system Python is easier and makes PyRosetta available at all times, though it makes upgrading PyRosetta more difficult and may require administrator access. Using an environment manager is more flexible and permits installation as a normal (non-admin) user, but requires more effort in understanding the system.
In addition to standard 'setup.py' package we provide Python wheel packages (use '.wheel' download links). Wheel package significantly reduce install time but requires `pip` to be present on your system. To install PyRosetta from wheel use `pip install pyrosetta-<version>.whl` command.
System-wide install (using system Python):
Download the appropriate version of PyRosetta from the links above.
Unpack the downloaded file to the location of your choice to create the PyRosetta directory.
(From a terminal/console window, you can unpack the archive using the command: tar -vjxf PyRosetta-<version>.tar.bz2.
For 'python' package from within the new PyRosetta directory, type cd setup && sudo python setup.py install into the command line to set up the PyRosetta.
For wheel package (recommended ) do: pip install <path-to-wheel-package>
Start Python.
In Python, you should be able to import the PyRosetta library with the command import pyrosetta; pyrosetta.init().
(If this step does not produce a complaint or error, your installation has been successful.)
Once you've installed PyRosetta, the downloaded PyRosetta archive file and the associated directory from steps 2 and 3 are no longer needed.
Before beginning, be sure that your desired environment manager (e.g. conda or virtualenv) is installed in your system. Please see tutorials for the corresponding environment manager to learn how to install and use them.
Download the appropriate version of PyRosetta from the links above.
Using the environment manager, create a new environment for this version of PyRosetta. (conda; virtualenv)
(The corresponding version of Python should be the only dependency required.)
Activate the new environment.
Unpack the downloaded file to the location of your choice to create the PyRosetta directory.
(From a terminal/console window, you can unpack the archive using the command: tar -vjxf PyRosetta-<version>.tar.bz2.
From within the new PyRosetta directory, type cd setup && python setup.py install into the command line to set up the PyRosetta.
Start Python.
In Python, you should be able to import the PyRosetta library with the command import pyrosetta; pyrosetta.init().
(If this step does not produce a complaint or error, your installation has been successful.)
Once you've installed PyRosetta, the downloaded PyRosetta archive file and the associated directory from steps 2 and 5 are no longer needed.
If you installed PyRosetta within an environment manager, remember to re-activate the corresponding PyRosetta environment each time you wish to use PyRosetta.
Follow the [Instructions for Installing PyRosetta on Windows 10]