The servers available for the protein-protein docking generate a thousand of the structures where they talk about the probability of finding structures in those 1000 structures. Is there any way I can refine my search so as to find the best and reliable model? These servers are based on ACE potential statistics, Space complementarity and electrostatics.
The first selection criterion would be the energy of the complex. the lowest energy complex is preferably considered.
The best method is to analyze the interactions in maximum possible number of complexes which leads to the identification of similar complexes with similar interactions. the maximum likely looking complex among them may be considered and also based on the energy.
Hi Nandakumar
Can you please suggest me some software or algorithm to find such similarities. Also about the minimum lowest energy of complex. I have the complex structures do you know any software or algorithm to calculate the minimum energy of the state.
I thank you for your kind help.
Dear Anupam,
U can get the minimum energy state of ur P-P complex by doing energy minimization or optimization. S/w 's like CHIMERA ,Argus Lab, Modeller can do this calculation.
PS. Based on the interaction (type of interactions and type of residues that interact) u can choose it.
The list of docked poses that the server gives you should be ranked according to some scoring function, and the best structure should be the top ranked one. This is what happens if you use, for instance, the SwarmDock server which I developed: http://bmm.cancerresearchuk.org/~SwarmDock/
If you want to re-rank your poses you already have, then there are many options available to you. I recently published a benchmark of many different publically available scoring functions for this purpose: http://www.biomedcentral.com/1471-2105/14/286/
One of the best methods was the latest version of ZRANK, which you can download for free.
As the next step, we usually do the energy scoring in one of the most frequently-used force fields (charmm or amber) for the selected, best poses, or all available poses (it doesn't take much time, so calculating interaction energy for 1000 poses is fine).
You should anyhow do at least some short Molecular Dynamics run for the selected, best pose just to see if the complex is stable over a period of time, so converting to amber or charmm would be needed anyhow. For free software I'd recommend the commonly-used namd (http://www.ks.uiuc.edu/Research/namd/) and standard charmm force field.
Just a word of caution: protein-protein docking is very tricky and finding the correct pose is many times very challenging.
HEX is the best easy to use free ware for protein-protein docking.
Discovery studio and schrodinger are the effective but commercial
You can calculate free Gibbs free energy or enthalpy using www.chemosophia.com on-line computations. Then you can choose the complex with the lowest Gibbs free energy or enthalpy. Just register at the web page, upload *.sdf file of "protein-protein" complex, select type of calculations "complex", find "Large organic and elementorganic molecular systems (up to 10 000 atoms)" service and start calculations.
Thank you all for your kind help.
I am trying to incorporate these suggestions.
Keep me posted if anything else can be done in order to enhance the accuracy of the results.
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology