AMBER Archive (2003)

Subject: AMBER: zero free energy change for nonbond leg of electrst. decoupling

From: Alfonso Garcia Sosa (atg21_at_hermes.cam.ac.uk)
Date: Tue Aug 05 2003 - 09:03:59 CDT


Dear amber list

 I have been running some free energy perturbations
(thermodynamic integration, dynamically modified windows) with
electrostatic decoupling in gibbs (that came with amber7) and
had no problems during the first leg, ielper = 1, that is,
change in electrostatics and nothing else. For the second leg
(change in vdW and other non-electrostatics, fixed electrostatics),
I set ielper = -1, and started a new run. The problem I have is that
the values for energy changes are all zero for the whole length of
the new run and also after completion. I have turned SHAKE
(ntc=2,ntf=2) on and off with no change, and also tried different
timesteps (dt=0.002 and dt=0.001), and restraint values.
I'm I running the calc. with the proper settings?
Thank you

input file (.in):

TI VdW, no elec.
 &cntrl
   ntb = 0, tempi = 300.0, temp0 = 300.0,
   dt = 0.002, nstmeq = 50, nstmul =50,
   ntc = 2, ntf = 2, intprt = 0, cut = 15,
   ntpr = 100, nrun = 1,
   isldyn = -2, ielper = -1, idifrg = 1,
   ntx = 5, NTR = 1, ivcap = 0, fcap = 10.0,
   ntxo = 1, ntwx = 500, ntwe = 500, ntt = 1,
   vlimit = 15.0, scnb = 2.0, scee = 1.2,
   idiel = 0, isande = 1, ntrx = 1,
   nstlim = 70000, iatcmp = 1, ntatdp = -1,
   ndmpmc = 100, IVEMAX = 0,
   IG = 71277, irest=1, ncorc = 0,
 &end
Hold the protein backbone
1.0
FIND
* C * *
* CA * *
* N * *
* O * *
SEARCH
RES 1 126
END
Hold the ions
0.7
RES 127 128
END
END

excerpts from output file (.out):
.
.
.
(snip)

NSTEP = 69900 TIME(PS) =139.800 TEMP(K) = 302.717 PRESS = 0.00
 Etotal = 1692.34570 Kinetic = 1442.83408 Potential = 249.51161
 Bond = 349.96634 Angle = 937.75345 Dihed = 1239.23327
 1-4 VdW = 441.60354 1-4 Elec= 1742.92467 Van d Waal = -760.95881
 Elect. = -3775.90399 H-bond = 0.00000 Constraint = 74.89314
 Ekcmt = 0.00000 VirT = 0.00000 Volume = 0.00000

Current Lambda = 0.399800
 Last F.E. update: Lambda = 0.499800 Step = 64000 Method = T.I.
 Accumulated "forward" quantities (Nonbond change)
    Lambda = 0.499800 F_energy = 0.00000
    Enthalpy = 0.00000 T*Entropy = 0.00000
    (Note: values are unchanged from last report)
 ------------------------------------------------------------------------------
 NO DYNAMICALLY MODIFIED WINDOWS COLLECTED SINCE LAST REPORT

      A V E R A G E S O V E R 70000 S T E P S

 NSTEP = 70000 TIME(PS) = 140.000 TEMP(K) = 299.919 PRESS = 0.00
 Etotal = 1686.60932 Kinetic = 1429.49756 Potential = 257.11176
 Bond = 348.58784 Angle = 957.07933 Dihed = 1244.51479
 1-4 VdW = 444.87314 1-4 Elec= 1752.52797 Van d Waal = -785.35587
 Elect. = -3788.03449 H-bond = 0.00000 Constraint = 82.91905
 Ekcmt = 0.00000 VirT = 0.00000 Volume = 0.00000
      R M S F L U C T U A T I O N S
 NSTEP = 70000 TIME(PS) = 140.000 TEMP(K) = 4.019 PRESS = 0.00
 Etotal = 20.42288 Kinetic = 19.15710 Potential = 28.93542
 Bond = 14.41945 Angle = 27.69824 Dihed = 12.76417
 1-4 VdW = 8.28349 1-4 Elec= 15.68998 Van d Waal = 16.39992
 Elect. = 24.99593 H-bond = 0.00000 Constraint = 6.39360
 Ekcmt = 0.00000 VirT = 0.00000 Volume = 0.00000
 Current Lambda = 0.399800
 Last F.E. update: Lambda = 0.499800 Step = 64000 Method = T.I.
 Accumulated "forward" quantities (Nonbond change)
    Lambda = 0.499800 F_energy = 0.00000
    Enthalpy = 0.00000 T*Entropy = 0.00000
   ELEC = 0.000 NONB = 0.000 14NB = 0.000
    14EL = 0.000 BADH = 0.000
 Quantities used in integration:
    <dV/d_lam> = 0.38978385E-15 <V(lam-dep.)> = 165.62357
    <V(lam-indep.)> = 108.33314 <V(lam-dep)*dV/d_lam> = 0.64658803E-13
    <V(lam-indep)*dV/d_lam> = 0.41822873E-13

------------------------------------------------------------------------------
 DYNAMICALLY MODIFIED WINDOW STATISTICS:
 Averages (Abs. values); Lambda Interval: 1.000000-->0.399800 NWINDOWS= 8
 delta(LAMBDA)= 0.0750250 dA/d(LAMBDA) [SLOPE]= 0.00000
 slope*delta(LAMBDA)= 0.00000 corr. coef.= 1.000000 pts for line= 5.00
 delA(for)-delA(rev)= 0.00000 multiplier= 1.00000
 delta(free energy): 0.00000 max. dA= 0.00000 min. dA= *************
 # Restarts due to AMXRST= 0 # Steps set to DLMIN= 0 DLMAX= 0
     Routine Sec %
     ---------------------------- | Pairlist 407.28 1.19 | Nonbond
31650.68 92.74 | Bond 37.73 0.11 | Angle 347.71 1.02 | Dihedral 1294.79
3.79 | PMF 0.46 0.00 | Constraint 124.92 0.37 | Other 263.16 0.77 |
Nonsetup ******************
      ---------------------------- | Total 34126.72 9.48 Hours | Setup
wallclock_t 0 seconds | Nonsetup wallclock_t 34127 seconds

I then restarted and final snippet is:
.
.
.
Current Lambda = 0.000000
 Last F.E. update: Lambda = 0.000000 Step = 500 Method = T.I.
 Accumulated "forward" quantities (Nonbond change)
    Lambda = 0.000000 F_energy = 0.00000
    Enthalpy = 0.00000 T*Entropy = 0.00000
    ELEC = 0.000 NONB = 0.000 14NB = 0.000
    14EL = 0.000 BADH = 0.000
 Quantities used in integration:
    <dV/d_lam> = 0.27667605E-15 <V(lam-dep.)> = 61.015842
    <V(lam-indep.)> = 131.87194 <V(lam-dep)*dV/d_lam> = 0.16973118E-13
    <V(lam-indep)*dV/d_lam> = 0.36595255E-13

------------------------------------------------------------------------------
 DYNAMICALLY MODIFIED WINDOW STATISTICS:
 Averages (Abs. values); Lambda Interval: 0.399800-->0.000000 NWINDOWS= 5
 delta(LAMBDA)= 0.1000000 dA/d(LAMBDA) [SLOPE]= 0.00000
 slope*delta(LAMBDA)= 0.00000 corr. coef.= 1.000000 pts for line= 8.00
 delA(for)-delA(rev)= 0.00000 multiplier= 1.00000
 delta(free energy): 0.00000 max. dA= 0.00000 min. dA= *************
 # Restarts due to AMXRST= 0 # Steps set to DLMIN= 0 DLMAX= 0

(ends)

Perhaps relevant?: I'm running mpi on a SunOS 5.8 Generic_108528-14 sun4u
sparc SUNW,Ultra-80 grid. Thanks for any help

--
Alfonso Garcia Sosa
PhD student
Dept. Pharmacology
Univ. of Cambridge, UK

----------------------------------------------------------------------- The AMBER Mail Reflector To post, send mail to amber_at_scripps.edu To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu