Parameters in the control file relevant for `flooding'

The following parameters in the control file ctl.lis are used for conformational flooding:

Switch for using flooding
Set this switch to TRUE in order to include a flooding potential into the force field. The flooding matrix flooding.lis, which can be generated using mkflood, is required. Translation/rotation correction should always be switched on. Appropriate flooding energy strengths (see below) must be given. Set this switch to FALSE to perform a conventional MD simulation. In this case all parameters given below are ignored, except for the Switch for using transl/rotation correction

Switch for using transl/rotation correction
If set to TRUE, this switch eliminates rigid body movements like translations and rotations of the selected atoms during a simulation. This feature is required for performing flooding simulations. Since the atom selection is given in the file flooding.lis, that file is always required if rigid body movements are to be eliminated. Rigid-body elimination is also necessary for the unperturbed MD-run required to generate a flooding matrix [step (4) above]. Don't be confused by the fact that here, of course, you will not have a flooding-matrix at hand. Instead, a dummy-file flooding.lis can be created using mkflood, in which only the atom selection is contained (see the sample application given below).

Switch for using adaptive flooding

THIS OPTION IS NOT YET IMPLEMENTED.

PLEASE SET THIS SWITCH TO ` FALSE'.

Adaptive flooding is an issue where indeed some knowledge on the theory section is beneficial. If you have not yet read that section, we recommend setting this switch to FALSE, in which case the Time constant for adaptive flooding is ignored. We emphasize, however, that there are cases in which conformational flooding will work much better when using adaptive flooding.

When performing a flooding simulation, two quantities can be used to characterize and to adjust the flooding strength: (a) the flooding energy $E_{\rm fl}$, which determines the `height' of the flooding potential $V_{\rm fl}$ and (b) the average value (a running average with an appropriate time constant) of the actual value of the flooding potential $V_{\rm fl}$, which is a function of configuration and, therefore, of time. It can be shown that this average is a good estimate for the flooding destabilization free energy $\Delta F$, which in turn determines the expected acceleration factor $\alpha\approx\exp(\Delta F/k_bT)$ via the Boltzmann factor. In contrast, the relation between the flooding energy $E_{\rm fl}$ and the expected acceleration is not quite clear and depends sensitively on the quality of the Gaussian configuration space density approximation.

Thus setting a definite value for the flooding energy $E_{\rm fl}$(switch: FALSE) makes it somewhat difficult to estimate its effect and typically entails a couple of trial-and-errors.

On the other hand, dynamically adjusting the flooding energy $E_{\rm fl}$such that a definite value for the destabilization free energy $\Delta F$(and, therefore, for the expected acceleration) is obtained (switch: TRUE) makes life much easier, but requires more experience in the interpretation of the development of the flooding energy in time.

Initial energy for flooding in kT
Initial flooding energy in units of thermal energy. Target temperature in Kelvin is used to translate that value into kcal/mol. Typical values are zero or small values, if increasing flooding strength is required, or a value equal to Final energy for flooding in kT, if constant strength is required. Depending of the Switch for using adaptive flooding, this energy is interpreted in two different ways: If the switch is set to FALSE, the energy denotes the flooding energy $E_{\rm fl}$; if set to TRUE, a target value for the actual averaged value (running average with time constant Time constant for adaptive flooding in s) of the flooding potential $V_{\rm fl}$ is given here.

Final energy for flooding in kT
Maximum flooding energy used in units of thermal energy. Target temperature in Kelvin is used to translate that value into kcal/mol. Flooding energy increase in kT/ps is used to determine how fast this maximum value is reached. Depending on the Switch for using adaptive flooding, this energy is interpreted in two different ways: If set to FALSE, the energy denotes the flooding energy $E_{\rm fl}$; if set to TRUE, a target value for the actual averaged value (running average with time constant Time constant for adaptive flooding in s) of the flooding potential $V_{\rm fl}$ is given here.

Flooding energy increase in kT/ps
The actual flooding energy used during the simulation is (Initial energy for flooding in kT) $+({\rm elapsed~time})*$(Flooding energy increase in kT/ps) as long as this expression is smaller than the Final energy for flooding in kT. Otherwise, the flooding energy is set to Final energy for flooding in kT. This allows (1) to have constant flooding energy (increase 0), (2) linear increase with time and (3) linear increase with time followed by a constant maximal value of the flooding energy.

Time constant for adaptive flooding in s
If Switch for using adaptive flooding is set TRUE, this time constant is used for averaging of the actual flooding potential in order to estimate the actual destabilization free energy $\Delta F$. That average is compared with the target flooding energy, and $E_{\rm fl}$ is adjusted dynamically such as to minimize the difference of both.