Computational Drug Discovery and Design

(backadmin) #1
MD simulations. It can be further used to populate the conforma-
tional ensemble of a multiconformational docking. The RC nonde-
pendent method group exploits the fact that the free energy is a
state function, so free energy differences can be calculated to pro-
vide binding thermodynamic quantities regardless of the path.
However, from such methods, we cannot extract kinetic informa-
tion from free energy estimation because they need the definition of
(un)binding pathway along which energy barriers and possible
intermediates can be determined. Interestingly, scaled MD and
similar smoothed approaches have been successfully employed to
predict residence time from the simulation time, even though they
are RC nondependent methods [42, 43]. The US is convenient to
extract both thermodynamic and kinetic information through the
construction of the potential of mean force (PMF, Fig.6)
corresponding to a projection of the free energy along the reaction
coordinate.

4 Notes



  1. In the absence of crystallographic structure or presence of gaps
    in the structure, homology modeling can be envisaged.

  2. There is no universal protocol for thermal equilibration. The
    proposed five-step protocol can be modified depending on the
    system.

  3. The choice of the method depends on what we want to simu-
    late but also on what we planned to analyze. For example, if we
    want to generate several diverse conformations of the protein as
    we did in our example using aMD, then a method using a
    reaction coordinate is not appropriate. We can use aMD simu-
    lation, REMD simulation, simulating annealing calculation
    (Table2) or Monte-Carlo methods. However, if the purpose
    is to analyze the stability of the conformations through the
    estimation of their free energy then aMD simulation is more
    suitable. Now, if the purpose is to sample the conformations of
    a protein between two states, then methods using reaction
    coordinate are more suitable because it allows guiding the
    sampling.

  4. When several methods are possible for a given study, an impor-
    tant reflex to have is to compare the performance of the meth-
    ods and the supported hardware.

  5. When using an enhanced molecular dynamics method, impor-
    tant preliminary analyzes must be made to ensure that the bias
    is not too strong and does not damage the system at the
    considered physiological temperature. We recommend to
    always assess the protein structure stability and fold analysis


422 Sonia Ziada et al.

Free download pdf