GeinDock Suite
Home › Which mode do I need?

Which mode do I need?

Four docking protocols, what each of them allows to move, what each asks you to supply, and how to choose between them.

GeinDock Suite runs four docking protocols. They share the same receptor, the same ligands and the same binding site; what differs is what is allowed to move, and what a pose has to satisfy before it is reported. You choose between them from Docking › Docking Mode, and the Docking Panel reflects the choice.

The docking mode submenu
The four protocols. Exactly one is active at any time, and the active one is ticked.

A short decision guide

Ask yourself If the answer is yes Mode
Do I just want to know how this compound sits in this pocket, and how well?Start here. It is the fastest, it is what the other three build on, and it is the right answer for most questions.Ligand docking
Does the pocket change shape when something binds, so a rigid receptor would refuse a compound that really fits?Let the side chains that line the pocket move during the search.Flexible docking
Do I already know something the pose must satisfy, such as a contact with a particular residue or a scaffold that must sit where a known compound sits?State that condition and have it enforced rather than hoping the search finds it.Constraint docking
Does the compound form a chemical bond to the protein rather than simply sitting in it?Anchor it to the residue it bonds to.Covalent docking

What each mode changes

Receptor Ligand Step 4 becomes What you supply
LigandRigidFlexible about its rotatable bondsSettingsSearch configuration
FlexibleSelected side chains moveFlexibleFlexible Side ChainsWhich residues may move
ConstraintRigidFlexibleConstraintsThe geometric conditions a pose must meet
CovalentThe anchoring residue movesBonded to the receptorCovalent AttachmentWhich residue atom and which ligand atom carry the bond
The step number never moves. Whichever mode you are in there are still six numbered steps and step 4 is still the fourth of them. Only its identity changes, so your mental map of the workflow survives a mode switch, and the panel keeps whatever you folded away.

Why three of the four have no settings page

Ligand docking exposes a settings dialog. The other three do not, and this is a design decision rather than an omission.

Each of the specialised protocols derives its own search parameters from the method it implements: the residue selection rules, the way a constraint is turned into something the search can act on, the anchor geometry of a covalent attachment, and the search parameters that go with each. Those derivations are the part of the protocol that was validated. Presenting them as fields you could edit would let a run be invalidated by a change that looked harmless at the time and was impossible to spot afterwards.

Nothing is hidden. Each mode's dialog shows the derived values, and the log records them with every run, so the configuration is always recoverable. It simply is not adjustable.

The verdict a run reports

The specialised modes ask for something specific, so they can report whether they got it. That verdict appears in the Result column of the Job Queue, in the pose table, and it travels with the saved results.

Verdict Meaning
All metEvery condition the protocol was asked to enforce was satisfied by the reported poses.
WarningThe run produced usable poses, but something about them deserves a look before you rely on the result.
Not metThe run executed correctly and produced poses, but none of them satisfied the condition you set.
Not met is not the same as failed. A run that prepared everything, searched properly and wrote nine poses, none of which reached the residue you constrained, is a run that worked and a criterion that was not met. The poses on disk are real evidence and are worth looking at, because they tell you what the compound does instead. The word failed is reserved for a step that genuinely did not execute.

Switching modes partway through

Switching modes keeps everything except step 4. The receptor, the ligands and the binding site stay exactly as they are, so you can dock the same system three ways without repeating any preparation. What you configured in one mode's step 4 is remembered for when you come back to it.

Run ligand docking first, whatever you intend to end up with. An ordinary run is fast and it tells you whether the receptor, the ligands and the box are sound. If it produces nothing sensible, the problem is in the setup, and no amount of flexibility or constraint will rescue it. Fix the setup first, then add the sophistication.