DockM8: an all-in-one open-source platform for consensus virtual screening in drug design

Abstract

Virtual screening remains a critical step in structure-based drug design, yet variability in docking algorithms and scoring functions often limits its reliability. To address this challenge, we introduce DockM8, an open-source platform for consensus virtual screening that integrates pocket detection, ligand preparation, docking, rescoring, and consensus ranking within a single modular workflow. DockM8 supports five docking engines, 17 pose-selection methods, 17 scoring functions (SFs), and five consensus methods, all accessible through both a graphical user interface (GUI) and a Python application programming interface (API). Systematic evaluation on the DEKOIS 2.0, DUD-E, and Lit-PCBA datasets yielded median relative enrichments of 100%, 82.59%, and 8.85%, respectively, frequently matching or outperforming state-of-the-art approaches. Our findings further reveal that no pose-selection or consensus strategy universally excels, emphasizing the need for tailored, target-specific workflows. DockM8 is freely available under the GNU General Public License at https://github.com/DrugBud-Suite/DockM8. Scientific contribution DockM8 is, to our knowledge, the most comprehensive open-source platform to integrate the complete consensus structure-based virtual screening (SBVS) pipeline within a single, graphical-interface-driven tool, yielding over 55 million configurable protocols, a breadth unmatched by previously published consensus docking software. Through systematic multi-benchmark evaluation, we demonstrate that no single pose-selection or consensus strategy generalizes across targets, motivating a target-specific workflow-selection paradigm in place of prevailing one-size-fits-all heuristics. By delivering this tunability through a graphical interface rather than scripts, DockM8 brings rigorous, reproducible consensus virtual screening within reach of non-specialist medicinal chemists.