RxDock is a fast and versatile open-source docking program that can be used to dock small molecules against proteins and nucleic acids. It is designed for high-throughput virtual screening (HTVS) campaigns and binding mode prediction studies. RxDock started in 2019 as a fork of rDock (whose development has stalled since 2014) with the goal of updating the code for running on modern computer systems (including supercomputers), enabling cross-platform usage (primarily targeting GNU/Linux, FreeBSD, macOS, Illumos, and Windows), improving the command-line and application programming interfaces, and implementing newer and better performing algorithms.
Define cavities using known binders or with user-supplied 3D coordinates. Allow -OH and -NH2 receptor side chains to rotate. Add explicit solvent molecules and structural waters. Supply pharmacophoric restraints as a bias to guide docking.
Pre-processing of input files
Define common ligand structure for performing tethered docking (requires Open Babel Python bindings). Sort, filter or split ligand files for facilitating parallelization. Find HTVS protocol for optimizing calculation time. Pre-calculate grids to decrease subsequent calculation times.
Post-processing and analysis of results
Summarize results in a tabular format. Sort, filter, merge or split results files. Calculate RMSD with a reference structure taking into account internal symmetries (requires Open Babel Python bindings).
Binding mode prediction
Predict how a ligand will bind to a given molecule. The ASTEX non-redundant test set for proteins and DOCK and rDock test sets for RNA have been used for validating and comparing rDock with other programs.
High-throughput virtual screening
Run for million of compounds in short time by exploiting the capabilities of computer calculation farms. Ease of parallelization in relatively unlimited CPUs to optimize HTVS running times. The DUD set has been used for validating rDock and comparing its performance to other reference docking programs.