Bonds cannot break or form in classical MD, so none of what you describe sounds realistic. Perhaps the structure is distorting some, which indicates poor ligand parametrization.

Bonds cannot break or form in classical MD, so none of what you describe sounds realistic. Perhaps the structure is distorting some, which indicates poor ligand parametrization.

I think you might need Good optimization and Parameterisation of the Ligand,As per Justin-He told that too many times-Bonds wont form or break during Molecular dynamics simulations!!

Dear Manoj,

I do not think that this kind of simulation reflects what is happening inside the body. It seems that you started with poor initial structure. I suggest that you remake the structure and make sure that all bonds and other parameters are normal.

Good luck,

Rafik

Thank you all for your answers. I prepared the ligand structure in maestro and ligand preparation was done by lligprep where structure was energy minimized. then i docked in Glide (Schrodinger). The detailed procedure of MD is as follows:

The simulations were performed for a time scale of 30 ns, with GROMOS96 43a2 force field using the GROMACS 5.0.4 package. The topology files necessary for the simulation were generated by using PRODRG. Initially, protonation states were added to the protein residues at a pH-7, and the protein−ligand complex was simulated with explicit simple point charge water model in a simulation box of 1.0 nm from the surface of the protein. The system was neutralized by adding five sodium ions and energy minimization was performed using steepest descent method until reaching a tolerance of 100 kJ/mol. The system was equilibrated with temperature up to 300 K using the Berendsen thermostat and pressure 1 bar using Parrinello-Rahman barostat. Finally, 30 ns MD simulations were performed. The time step for the simulations was 2 fs, and the coordinates were stored every 2 ps. The analysis of trajectory was performed using GROMACS 5.0.4 package and the plots were viewed using GRACE.

The lligand structure was exactly correct before and after docking also. Is there any problem occuring during ligand parametrization before MD?

Have to seen the RMSD of protein and ligand ,hydrogen bonds and other interactions,please check all those and conclude

Further,i think GROMOS96 43a2 is united atom ff  and poor topology next time please try with better ff like charmm or amber you can make ligand topology from server provided for each ff

There is nothing inherently wrong with GROMOS96 43A2, but PRODRG topologies are notoriously bad and require significant manual adjustment. To diagnose any issues, do something simple and fast - simulate the ligand in vacuo and/or in a box of water. If it distorts, you have a topology problem (which you do anyway if you're using a stock PRODRG topology without fixing it).