Beyond imaging, which is worthwhile investigating, if it is NOT that, 10ns is fast for it to distort badly, however some groups (like DESRES) have reported issues with some protein-DNA complexes. What I would do is enforce weak WC base pair restraints while the system "settles" (likely a microsecond or so - better 3-5 microseconds) and then see what happens when you turn them off. I would also do this with multiple replicates (different ion distributions or initial velocities for 3-5 independent simulations).
--tec3
p.s. note that DNA base pair fraying up to 1-2 base pairs is expected on the 1 microsecond timescale for isolated DNA duplexes (aka not in complex with protein) and deeper fraying on longer timescales. Also if you want me to eyeball the molecular graphics, send a PDB stripped of the water (tec3.utah.edu).
________________________________________
From: b.ercig.nki.nl <b.ercig.nki.nl>
Sent: Tuesday, February 15, 2022 4:26 PM
To: amber.ambermd.org
Subject: [AMBER] DNA breaking down during the simulation
Hello everyone,
I am currently simulating a system which consists of 2 proteins and 1 dsDNA in complex. I have made an production run of 10ns without any problems however, when I extend the simulation after 10ns, I see that the nucleotides in DNA strand is separating from the strand for some reason. I am not sure if I was able to explain the situation. But this did not seem to be a problem with starting structure to me, since the complex structure was intact during the minimization, heating etc…
I used Amber FF14SB forcefield together with BSC1 forcefield.
solvateBox com TIP3PBOX 10.0 (also tried octahedral box; did not work also)
and I neutralize the system using Na+; Cl- ions.
Below is the parameters being used in the production run:
imin=0,irest=1,ntx=5,
nstlim=5000000,dt=0.002,
ntc=2,ntf=2,ig=-1,
cut=8.0, ntb=2, ntp=1, taup=2.0,
ntpr=500, ntwx=500, ntwr = 500, ioutfm=1,
ntt=3, gamma_ln=2.0,
temp0=300.0,
Maybe I am missing something really obvious here, and I would be really happy to hear your feedbacks. Thanks all !
Kind regards,
Bogac Ercig
Netherlands Cancer Institute
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Feb 15 2022 - 17:30:02 PST
