Dear Krishnashish,

Could you please elaborate on "DNA flies off to Boundaries"!!

if you are using PBC and the solute moves to the edge of the box it is not a problem you can later center it to the simulation box but if you are talking about rapid denaturation of GQ..then there must a problem with the FF description of simulation system.

Please let us know what is the scenario in your case solute movement in the box or denaturaion of solute??

Best

Raghav

Thanks for the reply.

There is no issue with the solute. There are 24 DNA strands, that break apart and move to the edge of the rectangular box. This happens in less than 0.1 ps, and I suspect that it is due to the force field. Is there any better force field than amber10 for nucleic acids? And also I do not understand why using Amber instead of OpenMM, I don't have this issue?

I used PME for nonbonded interactions and cutoff=1 nm.

As stated above, this is a PBC issue. OpenMM does funny things with PBC, centering the biomolecule at the coordinate origin but the remaining solvent at the center of the box. You have to re-image the trajectory using whatever tools you have available to you. If it were a force field issue, you'd be getting massive energy spikes, crashes, etc.

Hi Justin, thanks for the reply.  Is there anything I have to do with the periodic boundary conditions? 

Do you have any recommendations for the force field especially for G-quadruplex DNA?

You just have to re-image your trajectory later for viewing purposes, but it doesn't affect the physics.

There is extensive GQ literature that I suggest you spend some time looking at. Recent AMBER DNA/RNA FF improvements are really critical. But there is an inherent problem that ion-ion repulsion among the monovalent ions that normally stabilize GQ is unavoidably too large in any additive FF. We're working on a polarizable model but it's not ready for public consumption yet. Look into the work by Sponer and co-workers; they've done some of the most extensive work in this area.

Thank you. 

There are 24 DNA strands!!

so that means you built you GQ-nanowires using smaller repeating unit or there are 24-nanowires in your simulation box? for the first case please have a look at you coordinate file and check if any similar primes are in close proximity from repeating units (3'-3' or 5'-5').

Since past several years amber99sb_bsc0 has been extensively used by simulation community for nucleic acids. and As mentioned by Justin, Jiri Sponer and coworker's contribution to nucleic acid simulations field is remarkable. 

Best

Dear Justin

What software do you suggest for re-imaging trajectory? Both VMD and pymol are unable to remap based on the boundary conditions. 

1) As suggested by Raghvendra Pratap Singh, you can try amber99sb_parmbsc0. Very good for nucleic acids.

2) You can try simulation with the command "warp all off" if you are simulation using NAMD. Basically try to switch off the wrapping up of solute molecules whenever the solute moves to the boundaries.