Recently, I have been dealing with symmetrical docking of protein complexes and I have a problem. Well, I have no problems with docking complexes with cyclic symmetry, it is double-walled symmetry that causes me problems. I use the SAM program and the HSYMDOCK server and try to generate a model of human insulin complex with pdb id: 3aiy, as a docking subunit I took a pdb file containing the coordinates of the A and B chain atoms. The generated model seems to be good, but when I compare it in PYMOL with the structure native, then RMSD reaches 18 A.
I have no idea how to deal with this problem, am I taking the wrong subunit to dock?
Have you actually looked at the superposition? 18Å is in the same ballpark as the distance between two monomers in the hexamer.
If you say the docked complex "looks good", might the high rmsd just be an artefact of the subunits not being compared in the same order? Since you use PyMOL to calculate the rmsd, you can look at the alignment object to see whether the pairing of atoms makes sense. Superimpose just one subunit and see whether additional subunits fall into the more or less correct position, disregarding chain labels. Look at whether you found the correct dimer interfaces (the main beta-chain - beta chain contacts where the V-shape formed by the beta strand and the long alpha helix of the two monomers pack form an antiparallel arrangement with the beta-strands of the two units forming an anti-parallel beta sheet), and whether the three dimers arrange correctly.
In nature, insulin hexamers are usually formed at the extremely high concentrations in the secretory storage granules, where the hexameters are stabilised by two Zn2+ ions, each coordinating with His residues from three different subunits, and dissociate as insulin is released from the cells.
Is there a specific reason you compare to this NMR structure rather than to the X-ray structures? Unluckely, I cannot currently access the publication associated with this particular structure
Best approach for generate a homomer is using glaxy homomer, but if you have heteromer use modeller if you have an homologous template is correct , and if not use CA restraint. 1 Angstrom is use frequently.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Atomic, Molecular and Optical Physics
- Atoms