Ultrafast Shape Recognition-based Protein Model Quality Assessment using Deep Learning
Saisai Guo and Jun Liu
College of Information Engineering
Zhejiang University of Technology, Hangzhou 310023, China
Email: [email protected], [email protected]
Guijun Zhang, Prof
College of Information Engineering
Zhejiang University of Technology, Hangzhou 310023, China
Email: [email protected]
- Python > 3.5
- PyTorch 1.3
- PyRosetta
- Tested on Ubuntu 20.04 LTS
DeepUMQA.py
arguments:
input path to input
output path to output (folder path, npz, or csv)
optional arguments:
-h, --help show this help message and exit
--pdb, -pdb Running on a single pdb
--csv, -csv Writing results to a csv file
--per_res_only, -pr Writing per-residue accuracy only
--leaveTempFile, -lt Leaving temporary files
--process PROCESS, -p PROCESS
--featurize, -f Running only the featurization part
--reprocess, -r Reprocessing all feature files
--verbose, -v Activating verbose flag
--ensemble, -e Running with ensembling of 4 models.
-
Predicting
# Running on a folder of pdbs python DeepUMQA.py -r -v input/ output/ # Running on a single pdb file python DeepUMQA.py -r -v --pdb pdbfile -
Feature extracting
python DeepUMQA.py --featurize input/ outputFea/
- Traning
python train.py models/
The executable software and the source code of DeepUMQA is distributed free of charge as it is to any non-commercial users. The authors hold no liabilities to the performance of the program.