HADDOCK2.X and its web server are now in maintenance mode with only parameters and topologies being updated when support for new molecules is added, in line with what is happening in HADDOCK3.
Development priorities based on user feedback
Work is ongoing on the implementation of the Deeprank module, a graph neural network for scoring protein–protein models using protein language model features.
DeepRank-GNN-esm is currently available only as a stand-alone tool. Original publication is available: DOI: 10.1093/bioadv/vbad191.
All related developments can be found in our GitHub repository searching with the label “m|deeprank”.
Work is ongoing on the automatic generation of topology and parameter files for small ligands, based on the PRODRG software.
Related issues on GitHub are – issue #1475 and #1481
Following the release of a Jupyter notebook for the antibody-antigen tutorial, we plan to port other available HADDOCK3 tutorials to Jupyter notebooks.
In HADDOCK2.X we have introduced coarse-graining based on the Martini v2.2 force field. This feature is currently not available in HADDOCK3 and is being reimplemented. We also plan to provide support for the latest Martini force field.
All related developments can be found in our GitHub repository, in the CG-HADDOCK branch.
Implemented features based on user feedback
Following user requirements, a protocol for antibody-antigen modelling, using AI-generated models as input for the docking was published: DOI:10.1093/bioinformatics/btae583.
A protocol for nanobody-antigen modelling, using AI-generated models as input for the docking, is available as a preprint: DOI:10.1101/2025.07.01.662355.
Tutorials on both antibody-antigen and nanobody-antigen modelling are available from the HADDOCK3 tutorial page. Additionally, antibody-antigen tutorial is available as a Jupyter notebook that can be directly launched on Google Colab or EGI: Jupyter Notebook.
Following user requirements, the support for various carbohydrates in HADDOCK3 was added via the following pull request.
A protocol for protein-glycan docking was published: DOI: 10.1021/acs.jcim.4c01372.
Finally, a protein-glycan modelling tutorial is available from the HADDOCK3 tutorial page.
Following user requirements, a Contactmap module was develop to analyse the intermolecular contacts made between the various interfaces, reporting on their nature (e.g. charged-charged, hydrogen bonds, hydrophobic) and visualising those in a matrix or circular plot form.
All related developments can be found in our GitHub repository searching with the label “m|Contacmap”
Following user requirements, HADDOCK3 now produces easy-to-read analysis reports with all cluster statistics and visualisation of results as done on the HADDOCK2 web server.
All related developments can be found in our GitHub repository searching with the label “analysis/postprocessing”
The new Alascan module enables an automated alanine scanning of all interface residues to dissect the contribution of each residue to the interaction energetics. Custom mutations to any other amino acid types are also supported.
All related developments can be found in our GitHub repository searching with the label “m|alascan”
