Abstract
Proteins and nucleic acids function by interacting with each other and with other molecules. Protein interactions with DNA are the basis of key cellular processes, such as replication, transcription, packaging, and DNA repair. To understand the mechanisms of these processes, it is important to know the structure of the molecular components and their interactions. Structures of protein complexes with nucleic acids are especially difficult to solve experimentally due to their greater instability. Thus, relatively few experimentally determined structures of such complexes are available. Computational modeling is essential for understanding biomolecular mechanisms and for the development of our ability to modulate them. Advances in the protein structure prediction field have shown that protein models can routinely reach unprecedented levels of near-experimental accuracy. In this context, modeling macromolecular interactions has become a focal point in structural biology. The Dockground project increases our knowledge of macromolecular interfaces and provides data resources for the development of techniques for structure-based modeling of macromolecular interactions. The resource, which contains protein-protein and protein-RNA complexes, has now been expanded to protein-DNA interactions. To advance the scope and functionality of the resource, the expansion includes new automatic weekly update procedures and user-friendly web interface for search, analysis and download of the protein-DNA complexes. The utility of the new release of the resource is illustrated by an example of benchmarking AlphaFold3 predictions. Dockground is publicly available at https://dockground.compbio.ku.edu.