broad-spectrum-asap-paper

Structure-based prediction of affinity across viral families.

Intro

The rapid emergence of viruses with pandemic and epidemic potential presents a continuous threat for public health worldwide. With the typical drug discovery pipeline taking an average of 5-10 years to reach the pre-clinical stage, there is an urgent need for new strategies to design broad-spectrum antivirals that can target multiple viral family members and variants of concern. We present a structure-based computational pipeline designed to identify and evaluate broad-spectrum inhibitors across viral family members for a given target.

Publication: bioRxiv

Contributors

• Maria A. Castellanos
• Alexander M. Payne

Contents

Scripts and input files for the paper on structure-based prediction of affinity across the coronavirus family.

• Run sequence search and alignment with drugforge-spectrum from drugforge
• Run protein folding with ColabFold and structure alignment with drugforge-spectrum
• Ligand transfer, docking and refinement with drugforge-docking
• Scoring of poses with drugforge-score

Installation

To install this repository, follow these steps:

1. Clone the repository, then enter the source tree:

git clone https://github.com/choderalab/broad-spectrum-asap-paper.git cd broad-spectrum-asap-paper 2. Install drugforge from source (conda-forge installation coming soon) git clone https://github.com/choderalab/drugforge.git cd drugforge 3. Install drugforge-spectrum and all required individual packages (e.g., spectrum) mamba env create -f devtools/conda-envs/drugforge-spectrum.yml conda activate spectrum pip install drugforge-spectrum 4. [OPTIONAL] Install AutoDock Vina dependencies (Requirements: pip install -U meeko mamba install adfr-suite -c hcc # Not available for newer versions of MacOS!
5. Install ColabFold. This can be done locally using localfold, or via Docker, following the instructions on the ColabFold repo. The example will assume the program is installed on a module colabfold/v1.5.2.

1. Install gnina. Also not on available in conda-forge but can be installed via a Docker image. gnina is used for scoring docked poses and it not strictly required (the alternative is AutoDock Vina), but it is more accurate.

License

• This software is licensed under the MIT license - a copy of this license is provided as SOFTWARE_LICENSE
• The data in this repository is made available under the Creative Commons CC0 (“No Rights Reserved”) License - a copy of this license is provided as DATA_LICENSE

Copyright

Copyright (c) 2025, Maria A. Castellanos