How to Use LightDock

LightDock is a docking framework built on Glowworm Swarm Optimization, a multi-agent search strategy that explores receptor surfaces and ligand orientations through interacting swarms. The method is designed for macromolecular docking problems where a single rigid-body optimization can miss plausible interfaces or alternative encounter complexes.

On Neurosnap, researchers upload an Input Receptor, an Input Ligand, choose the Docking Mode, and optionally enable Flexible Backbone for a less rigid search. Simulation Steps controls how long the swarm-based exploration runs. This setup is useful for protein-protein docking, peptide docking, and protein-RNA or protein-DNA interaction modeling when both partners are already available as structures.

Rather than treating docking as one best pose, LightDock is best used as a ranked hypothesis generator. Researchers can compare multiple candidate complexes, inspect which swarms produced them, and decide whether the interface geometry is good enough for refinement or experiment.

The key idea in LightDock is the glowworm algorithm: many locally communicating agents sample orientations and positions around the receptor, update an internal luciferin score according to local fitness, and migrate toward more promising regions of the search space. That multi-scale strategy helps explore complex interaction landscapes more broadly than a single deterministic local search.

On Neurosnap, Docking Mode matters because the geometric and scoring assumptions differ for protein-protein, protein-RNA, and protein-DNA complexes. Flexible Backbone is useful when local structural adjustment is expected near the interface, but rigid runs are often the better first pass for cleaner or larger inputs.

Results should be interpreted comparatively. Strong docking candidates are those that rank well, recur across the top solutions, and make structural sense in the viewer, not merely the pose with the numerically lowest score.

• Macromolecular docking range: LightDock supports protein, peptide, RNA, and DNA partners rather than only small-molecule docking.
• Exploratory search: Glowworm Swarm Optimization is built to search broad interaction landscapes with multiple concurrent swarms.
• Useful controls: Docking Mode, Simulation Steps, and Flexible Backbone let researchers choose how aggressive and how flexible the search should be.
• Ranked pose review: The viewer and score table make it straightforward to compare top complexes before refinement or experiment.