I wanted to calculate binding free energy calculation of my complex (protein +inhibitor). I have no resources to calculate for large number of frames (ie, 5000 frames from 100ns) hence I have chosen only 20 frames (ie, 5ns each). I am using g_mmpbsa tool in gromacs. Using less frames are reliable to calculate binding free energy calculations?
You do not need to calculate energies for large number of frames because binding energies likely to be highly correlated at the smaller time-gap. The auto-correlation time of binding energy was varied from 2 ps to 500 ps depending on the protein-ligand complex as shown in g_mmpbsa paper.
Time gap between frames may affect the average binding energy depending on the molecular system. In the g_mmpbsa paper, average binding energies were compared with two different time-gaps (10 and 500 ps), and energies were almost similar. Average binding energies were calculated using boot strapping method.
What is the grid resolution you are using to calculate polar solvation energy? You may speed up calculation by using coarser grid resolution. You may also compare average binding energy using time-gaps of 1 and 5 ns to monitor the impact of time-gap on average binding energy.
have you tried using the method of Umbrella Sampling or metadinâmica? I believe these techniques of fast convergence, can help you. Umbrella is implemented in GROMACS, and metadinâmica can be used with the PLUMED (http://www.plumed-code.org/).
It depends on how representative of "reality" the frames you choose are. If you happen to choose a few frames where the protein-ligand complex has entered an unusual conformation (e.g. a temporary salt bridge has formed, etc) then your energies can be affected significantly. Your best bet may be to use several segments of the trajectory and compare the ddG you get from each. For example, you can discard the first 1000 frames as equilibration, then take frames 1000-1020, 2000-2020, 3000-3020, and so on. Ideally the results from each segment should converge. Of course getting a converged answer and getting the right answer are not necessarily the same thing so hopefully you have some other results (from e.g. experiment) to compare to. Hope this helps,
I have one question regarding time frame in g_mmpbsa. How to chosse time frame what command or option i have to add in g_mmpbsa -f md_o_1nopbc.xtc -s md_0_1.tpr -n chainABCDE.ndx -i ../5_simulation_conf8/pbsa.mdp -b 18000 -e 20000 -pdie 2 -pbsa -decomp
@Akhil Similar to other Gromacs modules: see help for -dt, -b and -e
thank you for reply sir but
what is the use of -dt is it use to choose frame at particular time interval. well i tried but get the same result
What is time-interval in the trajectory? If time interval in input trajectory is large than "dt", this option will have no effect.
Thank you so much everyone for your kind suggestions. (As I was quite busy with other stuffs i could not reply you promptly).
You may find better parameters here:
Article Assessing the Performance of the MM/PBSA and MM/GBSA Methods...
Article Molecular dynamics investigation on a series of HIV protease...
Rajendra Kumar
I am using g_mmpbsa tool for calculation of polar energy .
But i am facing an error. i had run MD simulation for 100 ns and i have selected 2000 number of frames.
but when reading just went to the middle system Aborted. and shows following error. I don't know how to resolve this problem.
I am using this command to run calculation :
g_mmpbsa -f dt.xtc -n h2ab.ndx -s h2ab.tpr -i mmpbsa.mdp -nomme -pbsa -decomp
I have attached the files with this email please guide me.
Parsing input file pX35HEFA.in... Parsed input file. Got paths for 1 molecules Reading PQR-format atom data from pX35HEFA.pqr. NOsh_setupCalcMGAUTO: Error! Finest mesh has fallen off the coarser meshes! NOsh_setupCalcMGAUTO: difference in 1-direction = 0.0005 VASSERT: ASSERTION FAILURE! filename nosh.c, line 1769, (0) 2032 atoms Centered at (1.128e+02, 1.726e+02, 5.209e+01) Net charge 1.70e+01 e g_mmpbsa.sh: line 1: 15436 Done
15437 Aborted (core dumped)
- Badges
- Science topic