I have read a paper (link at the bottom) that uses MM-PBSA calculation for calculating the binding energy for their protein-ligand complexes. However, they only used the last 1ns of their trajectories (the total length is 20 ns) with a stride of 100 ps (10 snapshots only).
My question is, would not it be better to include the whole simulation in calculating the binding affinity? or in other words what is the optimal length of MDS that I should use in MM-PBSA calculation? and from where should I start?
Article In silico identification of inhibitors targeting N-Terminal ...
I believe that you can take a look at your interaction's behaviors during your frames to decide which frames do represent the interaction at its peak and which do that at its weakest point, then sum up the frames of each condition to know the time during which you will calculate, and here you go.
The MM/PBSA is a relative binding energy method and it calculate the average. I will take those frames which are well equilibrated and take at least 500 frames. I will consider entropy while calculation to predict binding energy more accurately. For more follow below articles.
Article A statistical analysis on the performance of MMPBSA versus a...
Article End-Point Binding Free Energy Calculation with MM/PBSA and M...
Check the convergence of your MD trajectory and also don't forget to do cluster analysis to make sure you cover the entire conformational space . My advise to do many 1 ns MD simulation after equilibration for better calculations(It is also helps you to estimate the error of your MMPBSA calculation)(And don't forget to use different seeds for you runs). AND also use quasi harmonic entropy for better results. Moreover, MMGBSA in some cases could be better than MMPBSA.
Good luck.
Ibrahim M. Ibrahim Ahmed Samir Mohamed Krishnendu Bera Victor Golyshev check our new tool, gmx_MMPBSA... really worth a try... there you can check for stability of the calculated energy and have a better idea of which interval to chose
https://valdes-tresanco-ms.github.io/gmx_MMPBSA/analyzer/#types-of-charts
gmx_MMPBSA is compatible with all GROMACS... Different calculations can be performed with this tool, including but not limited to:
-binding free energy calculations with different implicit solvent model (PB, GB, and RISM)
-alanine scanning
-energy decomposition
-entropy corrections
-QM/MMGBSA
check this summary in the documentation for more details
https://valdes-tresanco-ms.github.io/gmx_MMPBSA/getting-started/
don't hesitate to contact us if you need any help
best!
Mario E. Valdés-Tresanco Thank you sir for sharing this tool. Can it be used with NAMD or just GROMACS?
Ibrahim M. Ibrahim at the moment we only provide support for GROMACS files... working on including NAMD as well tough :)
- Badges
- Science topic