Dear Gromacs users,
I am trying to study the stability of a fibrillar structure in aqueous environment using CHARMM36 m force field and GROMACS 2018 simulation package. I performed two 100 ns simulations. I choose different box length for two simulations. But, simulation boxes were set up such that there was atleast minimum distance of 15 Å between any protein atom and a box edge. In first simulation, total number of atoms is 70,880 and in 2nd simulation, total number of atoms is 111,007.
After 100ns simulation, I removed periodicity in both cases using gmx trjconv and loaded in VMD. I saw, in smaller box, two layer are trying so detach each other and backbone RMSD is significantly high (~0.8 nm) and in larger box, structure is relatively more stable and backbone RMSD is ~0.58 nm.
Does I need to simulate in large water box? What is the reason?
Any kind of advice/ suggestions will be deeply appreciated.
Thanks in advance,
Snehasis Chatterjee
The issue is not the box (which is already larger than is likely necessary) rather you haven’t imaged the molecules correctly. Center on a single chain that resides at/near the center of the fibril.
Dear Prof. Lemkul,
Hope, this mail finds you well. Thank you very much for your valuable suggestion. I used following commands to remove periodicity.
gmx trjconv -f combine.xtc -s md.tpr -pbc whole -o protein_pbc_whole.xtc
gmx trjconv -f protein_pbc_whole.xtc -s md.tpr -pbc nojump -o protein_pbc_nojump.xtc
gmx trjconv -f protein_pbc_nojump.xtc -s md.tpr -pbc mol -center -o protein_remov_periodicity.xtc
I loaded the trajectory into VMD and I didn’t see any periodic error. Now, I removed periodicity according to instruction, mentioned in your previous mail. I didn’t see significant change in RMSD compared to using the previos commands.
That means, in smaller box, two layer are trying so detach each other and backbone RMSD is significantly high (~0.8 nm) and in larger box, structure is relatively more stable and backbone RMSD is ~0.58 nm. I am confused at this point.
Any kind of advice/ suggestions will be deeply appreciated.
Thanks in advance,
Snehasis Chatterjee
I would not call the difference between 0.6 nm and 0.8 nm “significant,” especially if you only have one replicate under each condition. The box size should not be causing this, likely just intrinsic variation in your simulations. Run a few more replicates.
Dear Prof. Lemkul,
Thank you very much for your kind help.
Regards,
Snehasis