I would appreciate people's thoughts on a methodology consideration.
We're docking several thousand small molecules at a specific location on the receptor. We're using Autodoc Vina to reduce the library to e.g., 20 compounds. Then, further reduction follows using generalised Born and surface area solvation (MM/GBSA), etc.
This is my concern. The crystal structure is a substrate-bound enzyme. Those that prepared the structure mutated Glu to Gln to prevent activation. This single-point mutation is far from where we are docking the compounds e.g., > 4 nanometres.
We are considering what our risks and limitation in our study are before running anything. Will an in silico modification (reverting from Gln to Glu) in the crystal structure affect calculation performed by Autodock (vina of AD4) if those changes are outside the grid box in which docking calculations are performed?
I understand the calculations performed by Adutodock/Vina are stochastic to a degree, so a direct comparison of how that mutation affects the results would be tough to produce. However, is there any merit in performing a before and after screening on the receptor, even if that change is far from the docking site?
Apparently, the mutated site you mentioned exerts an allosteric effect on the receptor. Do you have information on how this mutation affects the conformation of the ligand binding site?
You pointed out specifically the impact of mutation of a single amino acid. .i.e it affects the activity of the enzyme when bound to the substrate. To understand this, you'll have to see the effect of this mutation on the tertiary architecture of the protein. MD can provide you with a slight detail in the changes upon mutation. If there are changes in the pocket due to the mutation away from the pocket, your docking results might vary.
Performing docking analysis for a protein-ligand complex is not exactly going to give comprehensive explanation on an activity potential. Like the routine goes, you identify the activation site then build a pocket around it to predict a poses (Rigid or flexible) that the ligand can form to sit well in the site, then score them based on RMSD for affinity strength. If you notice this condition, only the activation site is essential for this two function (pose forming and scoring). So if your protein structure is to have a mutant amino acid residue outside this box it has no influence in the docking profiling. Nonetheless, docking is not exactly a functional evaluation because the several other biochemical and chemical conditions that is needed is not present, the algorithm only mimic for structural effect.
If you need to do more on mutant functional effect on disease condition then that's a different analysis to pick on.
- Badges
- Science topic