An end-to-end framework for the prediction of protein structure and fitness from single sequence
SPIRED-Fitness
git clone https://github.com/Gonglab-THU/SPIRED-Fitness.git
cd SPIRED-Fitness
Install Anaconda
/ Miniconda
software
Follow the steps below to install the Python package or run the command conda env create -f environment/spired_fitness.yml
conda create -n spired_fitness python=3.11
conda activate spired_fitness
conda install pytorch cpuonly -c pytorch
pip install click==8.1.7
pip install einops==0.7.0
pip install pandas==2.1.4
pip install biopython==1.82
conda env create -f environment/gdfold2.yml
:exclamation: install PyRosetta
at PyRosetta LICENSE
conda create -n gdfold2 python=3.11
conda activate gdfold2
conda install pytorch pytorch-cuda=11.8 -c pytorch -c nvidia
pip install biopython==1.82
SPIRED-Fitness
git clone https://github.com/Gonglab-THU/SPIRED-Fitness.git
cd SPIRED-Fitness
docker build -t spired -f ./environment/Dockerfile .
Note: Please ensure that the container has a minimum of 50 GB of memory to run the models. Also, remember to adjust the Docker memory setting accordingly.
docker run -it --name=spired --memory=50g spired /bin/bash
model
folder. If you install from docker, then the model parameters have been pre-downloaded.# run SPIRED
bash run_spired.sh -i example_spired/test.fasta -o example_spired
# run SPIRED-Fitness
bash run_spired_fitness.sh -i example_fitness/test.fasta -o example_fitness
# run SPIRED-Stab
bash run_spired_stab.sh -i example_stab/test.fasta -o example_stab
An end-to-end framework for the prediction of protein structure and fitness from single sequence