Date: Mon, 18 Feb 2019 11:29:31 +0000
Hello everybody,
I've a question regarding umbrella sampling (US) with distances and further
analysis with the WHAM program of Alan Grossfield. I ran 1D US simulations
of two a Tryptophan-small-molecule interaction, where I am pulling them
away in explicit solvent condition. Input file for a window at 4A distance
looks like:
-----------------------------------------------------
3ns NPT production for angstrom
&cntrl
imin = 0, ntx = 5, irest = 1,
ntpr = 10000, ntwr = 20000, ntwx = 20000,
ntf = 2, ntc = 2, cut = 10.0,
ntb = 2, nstlim = 1500000, dt = 0.002,
temp0 = 300, ntt = 3,
gamma_ln = 1.0,
ntp = 1, pres0 = 1.0, taup = 5.0,
nmropt = 1, ioutfm=0, iwrap=0
&end
&wt
type='DUMPFREQ', istep1=1000,
&end
&wt
type='END',
&end
DISANG=dist_04.0
DUMPAVE=dist_04.0.dat
-------------dist_04.0----------------------------------------------------
&rst
iat=-1,-1
igr1=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19
igr2=25,27,28,34,38,42
r1=1, r2=04.0, r3=04.0, r4=17,
rk2=1.0, rk3=1.0,
&end
-----------------
Then I am using WHAM to get PMF
wham 1 17 50 0.00000001 300 0 wham.in wham.out 10 9999
where wham.in file has entries for each 13 windows:
win-03.0/dist01_03.0.dat 3.00 2.00
...
win-15.0/dist01_15.0.dat 15.00 2.00
Histograms looks perfectly fine, but the PMF shows me a strange decrease
after 8A while being pulled away further. I used a TIP4PEWBOX extending 30A
from the solute to make sure that solute going of of the box is not causing
any problem. I have tried many more replicas and longer simulations to
ensure sampling but results are same as here.
I have no clue why PMF starts decreasing while two molecules are separated
by 8A and there is nothing else to interact other than water molecules.
There are several -OH in small-molecule but I believe they should have a
minima while interacting to each other. I am including histogram (upper
panel) and PMF (lower) from the simulation.
Can someone point-out the possible cause of such a behavior?
[image: image.png]
Regards,
Mish
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
(image/png attachment: image.png)
