adabmDCA 2.0 – a flexible but easy-to-use package for Direct Coupling Analysis.

adabmDCA is available in three languages: C++ (single-core CPU), Julia (multi-core CPU) and Python (GPU-oriented). Follow the instructions for installing the desired implementation.

Open a terminal and run

pip install adabmDCA

Clone the repository locally and then install the requirements and the package. In a terminal, run:

git clone https://github.com/spqb/adabmDCApy.git
cd adabmDCApy
pip install .

📦 The main repository of the implementation can be found at adabmDCApy.

After installing Julia on your system, you can install the package in one of the following ways:

Open a terminal in the desired folder, and run the following commands:

# Download scripts from Github
wget -O adabmDCA.sh https://raw.githubusercontent.com/spqb/adabmDCA.jl/refs/heads/main/adabmDCA.sh
wget -O execute.jl https://raw.githubusercontent.com/spqb/adabmDCA.jl/refs/heads/main/execute.jl
chmod +x adabmDCA.sh

# Install ArgParse and adabmDCA.jl from the GitHub repo
julia --eval 'using Pkg; Pkg.add("ArgParse"); Pkg.add(PackageSpec(url="https://github.com/spqb/adabmDCA.jl"))'

This will install all necessary dependencies and set up the package.

1. Open Julia and install the package by running:

   using Pkg
   Pkg.add(url="https://github.com/spqb/adabmDCA.jl")
   Pkg.add("ArgParse")
   

2. Download the files adabmDCA.sh and execute.jl into the same folder

   wget https://github.com/spqb/adabmDCA.jl/blob/main/install.sh
   wget https://github.com/spqb/adabmDCA.jl/blob/main/execute.jl
   

3. Make the script executable by opening a terminal in the folder and running:

   chmod +x adabmDCA.sh
   

This will set up the package for use.

📦 The main repository of the implementation can be found at adabmDCA.jl.

1. Clone the repository

   git clone https://github.com/spqb/adabmDCAc.git
   

2. In the src folder run

   make
   

3. It will generate the executable file adabmDCA. In the main folder run also chmod +x adabmDCA.sh to use the main script file. See

   ./adabmDCA --help
   

   for a complete list of features.

📦 The main repository of the implementation can be found at adabmDCAc.

This package presents a new version of adabmDCA. The package comes in three different languages: C++ (single-core CPU), Julia (multi-core CPU), and Python (GPU-oriented). They share the same front-end interface from the terminal allowing the user to install and use one of the three equivalent versions based on hardware or software constraints.

We provide three different training routines:

• bmDCA: Trains a fully-connected DCA model;
• eaDCA: Trains a DCA model on a sparse coupling network by progressively adding couplings during the training;
• edDCA: Starts from a trained bmDCA model and iteratively removes the less informative couplings until the target sparsity is reached.

Additionally, we provide several routines for sampling and analyzing the generated sequences once a DCA model is trained, for constructing and evaluating - according to a DCA model - a single mutant library from a given wild type, and finally, for computing the pairwise contact scores, in terms of average-product corrected Frobenius norms of the DCA couplings.

Check out the online Documentation and the Colab notebook

This package is open-sourced under the MIT License.

If you use this package in your research, please cite:

Rosset, L., Netti, R., Muntoni, A.P., Weigt, M., & Zamponi, F. (2024). adabmDCA 2.0: A flexible but easy-to-use package for Direct Coupling Analysis.

This work was developed in collaboration with Sorbonne Université, Sapienza Università di Roma, and Politecnico di Torino.