ProQ4:

Protein Quality Estimation.

https://arxiv.org/abs/1804.06281

Installation instructions:

pip3 install numpy Cython &&
pip3 install .

You will also need a deep learning backend compatible with Keras. We recommend Tensorflow:

pip3 install -U tensorflow

You will also need dssp to be installed and in the path (as well as the Bio.PDB module from biopython

Usage instructions

Inside Python:

import proq4

model = proq4.get_proq4()
ali = proq4.process_a3m('path/to/alignment.a3m')
# Also available: process_aln and process_fasta

pred_1 = proq4.predict(model, ali, 'path/to/pdb1')
pred_2 = proq4.predict(model, ali, 'path/to/pdb2')

Example

example file to run and predict values for lddt is in casp-run.py

This command should work

python3 proq4/casp-run.py -msa PL-pro.fasta.a3m -pdb C1904TS003_1.pdb -out C1904TS003_1.QA (the results is save in proq4.qa for comparison=