Dear Hungie:
6 kcals over 5.6 ns? That's awful! Just kidding.
You'd need 2.8 million steps of integration to get 5.6 ns, assuming 2 fs
step size.
Which means, roughly, 2 x 10^(-6) kcal/mol would be your typical error
margine per step. Given the total energy of your system is around -3.8 x
10^4, you got a relative error (per step) at around 0.5 x 10^(-10). In
comparison, a 64-bit double precision has a relative accuracy of about
10^(-11). This assumes that the systematic error is accumulative.
I thought this is a remarkably accurate simulation.
yong
> -----Original Message-----
> From: A. Hungie [mailto:hungie01.hotmail.com]
> Sent: Friday, May 09, 2003 9:47 AM
> To: amber.heimdal.compchem.ucsf.edu
> Subject: Etot is not constant in NVE ensemble
>
>
> Dear All,
>
> I have run MD simulation of 14 basepairs DNA (in water and
> nuetralized by
> Na+ ions) using protocol as in Amber tutorial. After doing 6 rounds of
> minimizations (reducing restrain energy from 25 to 0
> kcal/mol.A^2), I ran MD
> in NPT ensemble for 100 ps which I obtained canstant box size. Then I
> switched to NVE ensemble for 5.6 ns (800 ps for each run). I
> found that
> total energy is not constant, i.e. it increases about 6
> kcal/mol (from
> -37828 kacl/mol to -37822 kcal/mol) within 5.6 ns. I used
> Amber 6 on Linux
> cluster and input file as below.
>
> Is the obtained total energy normal and acceptable? Is it equilibrium?
>
> #my input file.
> ------------------------------------------
> &cntrl
> ntx = 7, irest = 1, nmropt = 0,
> imin = 0,
> maxcyc = 1000, ncyc = 5000,
> ntmin = 1, dx0 = 0.1, dxm = 0.5, drms
> = 0.0001,
>
> nstlim = 800000,
> dt = 0.001, ndfmin = 0, t = 0.0,
> timlim = 999999., ntcm = 1, nscm = 2500,
>
> ntpr = 1000, ntwr = 1000,
> ntwx = 1000, ntwv = 0, ntwe = 0,
> ntwprt = 0, ioutfm = 0, ntrx = 1, ntxo = 1,
>
> cut = 8.0, dielc = 1.0, nsnb = 10,
> scnb = 2.0, scee = 1.2, iwrap = 0,
>
> ntb = 1, ntp = 0, npscal = 1,
> pres0 = 1.0, comp = 44.6, taup = 2.0,
>
> ntc = 2, tol = 0.000001,
> ntf = 2,
>
> ibelly = 0, ntr = 0,
>
> temp0 = 300.0, tempi = 300.0,
> ig = 71277, heat = 0.0, tautp = 4.0,
> ntt = 0, vlimit = 20.0, dtemp = 1.0,
> &end
>
> &ewald
> a = 42.0151161, b = 43.0746222, c = 69.8275997,
> alpha = 90.000, beta = 90.000, gamma = 90.000,
> nfft1 = 45, nfft2 = 45, nfft3 = 72,
> order = 4, dsum_tol = 0.00001, eedmeth = 1,
> opt_infl = 1, vdwmeth = 1, use_pme = 1,
> frc_int = 0, nbflag = 1, skinnb = 1.0
> &end
> -----------------------------------
>
> Thank you very much in advance.
>
> Hungie
>
> _________________________________________________________________
> Add photos to your messages with MSN 8. Get 2 months FREE*.
> http://join.msn.com/?page=features/featuredemail
>
Received on Fri May 09 2003 - 16:53:01 PDT
