A protein–protein docking run can produce many candidate poses, and a broad search may take time while leaving you with a long list to inspect. If you already have clues about where the proteins interact, you can use them to focus the search. The key is to constrain the starting geometry thoughtfully—without treating a narrower search as proof that a predicted pose is correct.
This workflow uses the Hex Extension in SAMSON. It is especially useful when you have a plausible binding region and want to reduce the orientations considered during docking.
Start with structures you can trust
Before configuring the search, make sure both protein structures are suitable for docking. The SAMSON tutorial’s preparation steps include removing atoms with alternate locations, removing waters, monatomic ions and ligands where present, and adding hydrogens if they are missing. If hydrogens already reflect a chosen protonation state, do not modify them casually. Missing residues or heavy atoms may need to be addressed separately; the tutorial points to PDBFixer for this kind of repair.
In the Hex interface, assign one protein as the receptor and the other as the ligand. If you may want to repeat the run with different parameters, save the ligand’s initial conformation first. This gives you a reference starting state to return to.
Use known binding-site information to focus the search
Hex can search over rotational orientations. With the Range angles sampling method, receptor and ligand range angles define spherical cones around the intermolecular axis—the axis connecting the origins of the two proteins. A range angle of 180° leaves the full sphere available; a smaller angle restricts which orientations are sampled.
If you know the approximate binding site, first move and orient the ligand so it is close to that site. Then set the receptor and ligand range angles in Advanced parameters. The tutorial demonstrates 45° for each as a way to loosely constrain the search around the starting orientation. These values are an example, not a universal setting: the useful range depends on how confidently you know the interface and how well the proteins are initially oriented.
You can also limit the ligand’s twist rotation about the intermolecular axis. Constraints can reduce the search space, but they also exclude orientations outside the allowed region. If the binding site is uncertain, overly tight limits could prevent relevant poses from being sampled. Treat the constraints as a way to express what you know—not as a substitute for testing alternative assumptions.
Balance speed and search detail
Hex offers a fast, lower-resolution Steric scan followed by a finer Final search. The first phase helps cover orientations coarsely; the final phase evaluates the retained candidates with smaller distance increments. The documentation notes that a Steric scan order of 16–18 is sufficient in most cases, and gives combinations such as N=16 for the scan with N=25 for the final search, or N=20 followed by N=30, as configurations that work well in many cases.
Final-search order is a trade-off. The default N=25 is often sufficient, while N=30 is recommended for high-resolution crystal structures when binding is expected to involve small conformational changes. For model-built structures or structures expected to be more flexible, the tutorial recommends N=25. A higher setting can take longer, so choose it in light of both the structure quality and the question you are asking.
Review poses as candidates, not answers
After docking, Hex clusters modes so you can inspect the best-scoring mode in each cluster or view all modes. Selecting a result restores that ligand conformation; you can also create a conformation in the document for further work. Use visual inspection and interaction analysis to compare candidates. SAMSON tools and Extensions can help you examine contacts, hydrogen bonds, surrounding residues, and other properties.
A docking score or a restricted search alone does not establish that a pose is biologically correct. The practical goal is to create a manageable, well-documented set of candidates for follow-up analysis—and to keep track of the assumptions that shaped the search.
For the full workflow, parameter details, and example results, see the Protein docking with Hex tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.