RoseTTAFold-local-server

It is a simple local server of RoseTTAFold or RoseTTAFold2. Ubuntu ≥ 20.04 is recommended. Remember change the folder path in the code before running it.

Environment:

python = 3.8.10
pandas
remi

The environment required pandas and remi based on python3. We recommend conda environment of RoseTTAFold or RoseTTAFold2. Just need install remi by conda install -c conda-forge remi. Then conda activate RoseTTAFold or conda activate RF2 and python Code/web.py. The webpage is start in http://0.0.0.0:46429.

You can install from requirement file:

conda install --yes --file RF_requirements.txt`    # RoseTTAFold
conda install --yes --file RF2_requirements.txt`   # RoseTTAFold2

Or, you can also create the environment by conda command:

conda env create -f RF_environment.yaml    # RoseTTAFold
conda env create -f RF2_environment.yaml   # RoseTTAFold2

Then

conda activate RFlocalserver   # RoseTTAFold
conda activate RF2             # RoseTTAFold2

Lastly, set up the server:

python Code/web.py

It's important to note that, background_service.py need RoseTTAFold environment:

conda activate RoseTTAFold   # RoseTTAFold
conda activate RF2           # RoseTTAFold2
python Code/background_service.py`

Context:

There are 2 parts: codes and running data.

Codes include 3 files:

Code/web.py: The website of the server. It will receive the amino acid sequence and write to apply.csv. The conda or python environment need remi library.

Code/background_server.py: The background running program of the server of RoseTTAFold. When apply.csv is not empty, it will create the run indicator file test.fa and start calculation. When finish indicator file model5.pdb exist, it will shear the imformation from apply.csv to result.csv. The results will be compressed into a tar.gz file and transferred to the result folder for web.py retrieval. Remember activate RoseTTAFold environment of conda before running it. The environment could refer to the github of RoseTTAFold.

Code/background_server_rf2.py: The background running program of the server of RoseTTAFold2. When apply.csv is not empty, it will create the run indicator file test.fa and start calculation. When finish indicator file model5.pdb exist, it will shear the imformation from apply.csv to result.csv. The results will be compressed into a tar.gz file and transferred to the result folder for web.py retrieval. Remember activate RoseTTAFold environment of conda before running it. The environment could refer to the github of RoseTTAFold.

Code/apply.sh The running file of RoseTTAFold. It will execute the RoseTTAFold run command and place the completed file in a specific location. We prefer pyrosetta than e2e.

Code/apply_rf2.sh The running file of RoseTTAFold2. It will execute the RoseTTAFold run command and place the completed file in a specific location. We prefer pyrosetta than e2e.

Running data includes 3 files and 1 folder:

RunningData/running folder include the calculating amino acid chain. After calculating, all the files in the folder will be compressed and move to specific folder.

RunningData/apply.csv includes the applied mission from web. When calculation finishes, the first line would be move to result.csv.

RunningData/result.csv records completed calculations. It includes the position of compressed file in specific folder.

Citing this work

Fei Liu, Xiangkang Jiang, Jingyuan Yang, Jiawei Tao, Mao Zhang, A chronotherapeutics-applicable multi-target therapeutics based on AI: Example of therapeutic hypothermia, Briefings in Bioinformatics, 2022;, bbac365, https://doi.org/10.1093/bib/bbac365

M. Baek, et al., Accurate prediction of protein structures and interactions using a three-track neural network, Science (2021).

I.R. Humphreys, J. Pei, M. Baek, A. Krishnakumar, et al, Computed structures of core eukaryotic protein complexes, Science (2021).

License

The license of the project is MPL-2.0.

Screenshot