Hi everybody,
I'm using Autodock4.2 and I have taken these steps respectively. Can anyone assure me that I have been on the right path? I'm concerning about "spacing" and the accuracy of each step, especially the big spacing and box size in the first step.
-step1 (whole molecule):
NELEMENTS=108*94*126 & Spacing=0.755
-step2 (focusing on docked part from step1):
NELEMENTS=126*126*126 & Spacing=0.316
-Step3 (set flexible amino acid at the binding site from step2): NELEMENTS=106*84*80 & Spacing=0.375
Thanks
Hi Homa,
Your first step is blind docking. So, I believe you look into the histogram and select the pocket i.e step2. If all the docking conformation is in one pocket then step2 correct otherwise need to consider other binding pockets also. Your step3 is correct.
Please go through this paper
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0075182
Thanks for the all answers
Prashantaha, I didn't find any certain binding site for this protein. What is the problem with blind docking if the software has this capability?
Dear Avanthika;
Actually I set a smaller box in the last two steps ,because the box size is a function of spacing value. There is a direct correlation between them when u are using a constant box size.
Thanks
Dear Homa,
Sure, the software has the "capacity" to do it, in the sense that the program will run with the parameters you provide him with. This does not guarantee the quality of the output. The larger box size and larger step will be at the expense of accuracy. A difference of 0.75A can mean a lot when a H-bond is concerned. As example, I have calculated and visualized (at -0.8 kcal/mole value) the OA maps for the same binding site, with 0.25A and 0.75A steps. There are clearly some differences. As a result, it might be more difficult for a docked compound to pick up the right interactions in your site.
Nonetheless, your approach remains quite correct, as you further "refine" your findings from step 1 with the more accurate settings. If the program is able to pick up the right binding site for your compounds at step 1, this is indeed what you should do.
If your compounds are congeneric, they probably bind to the same site. If so, the best would be to run a first "site identification" study with your larger grid box and a limited number of compounds (so, step 1). Then directly dock all your compounds in the identified binding site, with more accurate settings.
Ideally, you should obviously validate your hypothesis regarding the binding site at some point (agreement between SAR and docking; mutagenesis; whatever...).
Dear Sandro, although I'm not working with a library, I got some valuable advices.
Thanks Avanthika! I will follow your suggestion.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology