Fast and easy contact prediction.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Fast, accurate, and hassle-free contact prediction.

Installation instructions:

Download the tarball from the releases tab

pip3 install numpy Cython pythran &&
pip3 install pconsc4-0.2.tar.gz

NB: the trained model is a bit over Github's limit, so they cannot be checked in the repo, hence the need for the tarball.

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

pip3 install -U tensorflow

Usage instructions

Inside Python:

import pconsc4

model = pconsc4.get_pconsc4()

pred_1 = pconsc4.predict(model, 'path/to/alignment1')
pred_2 = pconsc4.predict(model, 'path/to/alignment2')

# Show pred_1 on the screen:

import matplotlib.pyplot as plt 

The program accepts alignments in .fasta, .a3m, or .aln, without line wrapping.

We also provide a function to format the output in CASP format:

# Save in CASP format:
from pconsc4.utils import format_contacts_casp
print(format_contacts_casp(pred_2['cmap'], seq_2, min_sep=5))


If pyGaussDCA fails to install with template errors, upgrade your compiler. GCC 5 and higher is known to work.