Dear Amber Users,
Every few years I try to revisit the implicit solvent MD for use with our fairly large RNA & DNA-based
nanostructures (300+ nucleotides). Think of a cube with edges made of A-form (RNA) or B-form (DNA)
helices and linked in the corners with single-stranded bridges. All helical edges are fully paired by design.
Some of the edges contain the 5'-3' breaks in the middle (i.e. one side is connected with the standard
backbone, the other held in place by stacking interactions).
I am evaluating a Langevin protocol in Amber 14 (ff14SB, igb=1, ntt=3, gamma_ln=1.0,
offset=? (0.09 or 0.13) and saltcon=1.0) for such cubes (PBradii mbondi used in Leap).
DNA cubes experience distortions larger than in the explicit solvent simulations, but generally
maintain structural integrity, whereas RNA-base cubes show a tendency to lose the stacking
in the middle of the helices across the 5' - 3' breaks, and I see a rotation of the initially stacked
nucleotides away from each other (up to 180'). G-C|G-C or G-C|C-G combinations hold with some
distortions, but G-C|A-U or G-C|U-A tend to "twist open" (| - indicates backbone connection).
Questions:
1) Are there any new options/parameters for Implicit Solvent MD - something equivalent to the
Liu ae al. (A Novel Implicit Solvent Model...), Biophys J, 2013 implemented in an unfortunately
named STINKER?
2) I've been relying on the default offset (0.09) in the runs so far. I am evaluating 0.13. Should it
help with the described problem in RNA?
3) Setting gbsa=1 slows down the simulation speed well below that of the explicit solvent MD.
Is it expected or is it a sign of something seriously wrong? A flexing object like a cube may
experience more surface area changes than a simple duplex, and the question in my mind is if
ignoring it (gbsa=0) is justified. The calculated values appear to be ~ 0.5% of the total Energy.
4) Are other parameters more important than those mentioned above? I am using dt=0.002, for example.
If you have any suggestions or comments, I would appreciate them very much.
Best regards, Voytek Kasprzak
Wojciech (Voytek) Kasprzak (Contractor)
Analyst Programmer,
Basic Science Program,
Leidos Biomedical Research, Inc.
Frederick National Laboratory for Cancer Research (FNLCR)
Frederick, MD 21702
(301) 846 5537
http://binkley2.ncifcrf.gov/users/kasprzak
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jun 29 2016 - 07:30:06 PDT
