Structure prediction of alternative protein conformations
Cfold is a structure prediction network similar to AlphaFold2 that is trained on a conformational split of the PDB. Cfold is designed for predicting alternative conformations of protein structures. Read more about it in the paper here
AlphaFold2 is available under the Apache License, Version 2.0 and so is Cfold, which is a derivative thereof. The Cfold parameters are made available under the terms of the CC BY 4.0 license.
You may not use these files except in compliance with the licenses.
The entire installation takes <1 hour on a standard computer.
The runtime will depend on the GPU you have available, the size of the protein
you are predicting and the number of samples taken. On an NVIDIA A100 GPU, the
prediction time is a few minutes per sample for a protein of a few hundred amino acids.
We assume you have CUDA12. For CUDA11, you will have to change the installation of some packages. \
First install miniconda, see: https://docs.conda.io/projects/miniconda/en/latest/miniconda-other-installer-links.html
bash install_dependencies.sh
bash predict.sh