My question is not how to choose the best binding pose per se, but do I need to based on any criteria?
For example, If I find that the best pose for my most active ligand is the pose with minimal binding energy. What should I do to analyze my other ligands? Look only for the pose with the minimal binding energy?
I'm asking this because I screened a library of 20 molecules with known activity. For the most active I selected the pose with the minimal binding energy to analyze. But my second best molecule (I already know it is active experimentally) only showed a promising complex after 15 binding poses (I calculated 20 poses for each ligand).
In case this helps, I used autodock vina.
Dear Marcos,
I have a doubt, What factors of your second best molecule conformations are you considering to think that the #15 pose is a promising binding mode? I suppose, you are taking into account structural properties i.e. the location in the binding site and orientation of functional groups perhaps.
The energetic is a quantitative and very important variable for the proposal of a binding mode but it is not the unique. You can use quantitative and qualitative properties based on experimental and bibliographic data (i.e. formation of a hydrogen bond, location of a molecule substructure in a specific pocket, orientation of functional groups…).
In this sense, Ali et al. 2008 (attached reference) propose a quantitative / qualitative methodology to select the most probable conformation of the complexes given by AutoDock. They first chosen the conformation with the lowest final docked energy as the starting point and analysed it qualitatively based on the location/orientation of the inhibitor in relation to the intact ligand.
On the other hand, perhaps, 20 conformations generated could not be enough to explore the neccesary conformational space.
I think you can design an "ad hoc" method based on energy and qualitative parameters which can explain the activity properties.
Hoping this can be useful.
Best
Miguel
http://www.sciencedirect.com/science/article/pii/S0223523407003911
Miguel,
I'm considering the binding mode and interactions with the enzyme. It's similar to those observed with other active molecules. The fact is, it's one of my last poses.
I guess it's worth trying to increase the number of conformations as you mentioned, but Vina doesnt seem to increase to more than 20 conformations (I'm checking it right now).
Your answer clarified some things to me. I'll take a look at that paper
Thank you!
i think you should check it for 100 runs by auto dock 4.2 or new version then you can easily select your best ligand
- Badges
- Science topic
- Similar topics
- Biological Science
- Neuroscience