Installation Guide¶
Using the Apptainer Image¶
If possible, we highly recommend using the provided apptainer image which can be obtained via:
curl -O https://files.ipd.uw.edu/pub/2025_RFDpoly/SE3nv.sif
Here is an example for how to use this image to run RFDpoly:
apptainer run --nv /path/to/SE3nv.sif /path/to/RFDpoly/rf_diffusion/run_inference.py --config-name=multi_polymer \
diffuser.T=50 \
inference.ckpt_path=/path/to/train_session2024-07-08_1720455712_BFF_3.00.pt \
inference.num_designs=1 \
contigmap.contigs=[\'33\ 33\ 75\'] \
contigmap.polymer_chains=[\'dna\',\'rna\',\'protein\'] \
inference.output_prefix=/path/to/your/output/directory/basic_uncond_test01
If you see an error when attempting to run this related to a lack of an input file, try adding this option:
inference.input_pdb=/path/to/RFDpoly/rf_diffusion/test_data/DBP035.pdb
Creating a Conda Environment¶
Note: Use pip 26.2.2 or newer. Older pip releases still trust the withdrawn emSign root CA certificate, so upgrade pip in the new environment before running
pip install, as shown below.
Linux:¶
If the apptainer image is not compatible with your system
or you are doing development with RFDpoly, you can create
a conda or mamba environment with a provided environment.yml file:
conda env create -f /path/to/RFDpoly/rf_diffusion/environment/environment.yml
conda activate RFDpoly_env
python -m pip install --upgrade pip # pip 26.2.2 or newer
pip install /path/to/RFDpoly/rf_diffusion/RF2-allatom/rf2aa/SE3Transformer
The last command installs a slightly modified version of
se3-transformer. If installing this package does not work,
you can try installing se3-transformer-pytorch instead,
but we do not guarantee that your results will be exactly
the same.
You can test your installation by running:
python /path/to/RFDpoly/rf_diffusion/run_inference.py \
--config-name=multi_polymer \
diffuser.T=50 \
inference.ckpt_path=/path/to/RFDpoly/weights/train_session2024-07-08_1720455712_BFF_3.00.pt \
inference.num_designs=1 \
contigmap.contigs=[\'33\ 33\ 75\'] \
contigmap.polymer_chains=[\'dna\',\'rna\',\'protein\'] \
inference.output_prefix=./basic_uncond_test01
If using the environment.yml file does not work, please
create an issue. Or, if you modified the .yml file to
work for your system, make a pull request to add it to
the rf_diffusion/environment folder.
MacOS¶
If the apptainer image is not compatible with your system
or you are doing development with RFDpoly, you can create
a conda or mamba environment with a provided macos_environment.yml file:
conda env create -f /path/to/RFDpoly/rf_diffusion/environment/macos_environment.yml
conda activate RFDpoly_env
python -m pip install --upgrade pip # pip 26.2.2 or newer
pip install /path/to/RFDpoly/rf_diffusion/RF2-allatom/rf2aa/SE3Transformer
The last command installs a slightly modified version of
se3-transformer. If installing this package does not work,
you can try installing se3-transformer-pytorch instead,
but we do not guarantee that your results will be exactly
the same.
You can test your installation by running:
python /path/to/RFDpoly/rf_diffusion/run_inference.py \
--config-name=multi_polymer \
diffuser.T=50 \
inference.ckpt_path=/path/to/RFDpoly/weights/train_session2024-07-08_1720455712_BFF_3.00.pt \
inference.num_designs=1 \
contigmap.contigs=[\'33\ 33\ 75\'] \
contigmap.polymer_chains=[\'dna\',\'rna\',\'protein\'] \
inference.output_prefix=./basic_uncond_test01
If using the macos_environment.yml file does not work,
please create an issue. Or, if you modified the .yml file
to work for your system, make a pull request to add it to
the rf_diffusion/environment folder.
Did you find what you were looking for?
Thanks for your feedback!