Recently I used the website "Hex Protein Docking Server". But I am suspicious whether the results are reliable or not . Also can it consider all types of binding (intercalating and groove binding) that depends on the drug? I need your help to clarify my doubt.
Reliability of a docking software is generally checked by testing its ability to reproduce crystallographic or experimental pose as pose with best predicted binding free energy value. By that, what I meant is that you should take a structural model in complex with some inhibitor. From your description it seems that you are doing DNA-Inhibitor interaction study (If that is correct, then take some DNA-Inhibitor complex). separately save the inhibitor and target molecule in two separate files. Now use the online software and dock this inhibitor molecule back into its receptor and obtain the complex. Now compare the predicted pose to that of crystallographic pose. If you get similar pose in prediction as compared to experimental pose, it means that you can trust this algorithm for your test compounds. Usually structural similarity is expressed in terms of RMSD values. So, you need to calculate RMSD between the predicted pose and experimental pose. If it is below 2 angstrom, you can rely on software. Ideally RMSD value should be below 1. So far as i know, Hex do not use ligand flexibility during docking runs (I.e it performs rigid body docking). Therefore it is possible that you might predict binding pose nearly similar to crystallographic pose. But you will not be able to explore ligand's alternative confirmations in receptors active site, owing to lower level of conformational space being explored by rigid body docking schemes. Rather, I would recommend you to use AutoDock for docking purpose.
Best luck for your docking exploration
Hi All. I am also using HexDock using RNA as the receptor and protein as ligand. I just wanted to know whether HexDock is an appropriate server for studying RNA-protein interaction? Please suggest.
- Badges
- Science topic