About the well-tempered metadynamics convergence..

From: zeynab hosseini (hosseinizeynab93_at_gmail.com)
Date: Thu Dec 10 2020 - 12:49:14 CST

Dear All,

I have performed 500ns of well-tempered metadynamics (wt-mtd) simulation to
calculate the potential of mean force (PMF) over the adsorption of a small
molecule on an infinite cabon nanotube surface (NAMD 2.13b1). The CNT is
oriented along the z-axis and the reaction coordinate is the radial
distance of molecule from the CNT central axis. The lowerboundary and
upperboundary are considered to be 4A and 30.5A, respectively. I have
plotted the PMF versus radial distance over 500ns (black curve), 496ns
(red), 492ns (green), 488ns (blue) and 484ns (yellow) of wt-mtd simulation (
link
<https://drive.google.com/file/d/1qaJmUTdVBsy6iEZoXTcR0W7Aj9u_uVp5/view?usp=sharing>).
The Jacobian term of the PMF (-kB T ln(r)) is included in the plot, too.

I wonder if I can conclude the PMF is converged after 500ns. After about
450ns, the PMF plot does not show any significant change from the plot I've
already attached. If so, then is that correct to calculate the average of
the five mentioned PMF curves as the final PMF, along with the standard
deviation to be considered as the error bar?

Is there any way to get more accurate PMF, e.g. all five curves overlap
each other entirley? Should I change the reaction coordinate or is there
any other way?

Is there anyway to get faster converngence? e.g. much shorter than 500ns.
For example to reduce the upperboundary from 30.5A to 20A? I'm not sure if
it affect the convergence positively.

Since the CNT and water box are symmetrical (CNT+molecule are solvated in a
water box of 70A × 70A × 24A), I wonder if I can confine the molecule
movements into a specific region, in order to get faster convergence. I
mean, e.g., if the CNT length is 20A (-10A < z < 10A), we apply restraints
to limit the molecule movements only in a region with -5A < z < 5A ? Does
it help to converge faster without affecting the physics of the system?

I would really appreciate if anybody guides me..

All the Best,
Zeynab

The relevant part of the metadynamics configuration file:
colvarsTrajFrequency 500
colvarsRestartFrequency 1000

colvar {
   name Translocation

   upperboundary 30.5
   lowerboundary 4.0
   upperWallConstant 100.0
   lowerWallConstant 100.0
   width 0.125331414

   distanceXY {
      main {
         atomnumbers { 165 }
      }
      ref {
         atomnumbers {
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
160
         }
      }
      axis ( 0.0, 0.0, 1.0 )
   }
}

metadynamics {
   name meta-distance
   colvars Translocation
   hillWeight 0.01
   newHillFrequency 250
   dumpFreeEnergyFile yes
   writeHillsTrajectory on
   hillwidth 1.253314137
   wellTempered on
 # bias factor=6
   biasTemperature 1550

}

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:15 CST