assemblies and transmembrane proteins makes the consideration of
the flexibility role more difficult to apprehend.
From the theoretical point of view, classical molecular dynamics
(cMD) simulation method is nowadays an appropriate way to study
the dynamics behavior of biological entities, even for large macro-
molecular systems. Starting from an experimentally determined
structure or from a computationally built model (comparative
modeling protocol), cMD simulations predict atomic motions of
a molecular system as a function of time. Due to the presence of
high energetic barriers on the potential energy surface of the sys-
tem, and the random nature of molecular dynamics simulation,
biological processes are stochastic events that usually occur on the
microsecond to millisecond or even longer timescales. Even with a
great improvement in computing hardware, cMD simulations are
limited to timescales of about several hundreds of nanoseconds,
making it unsuitable to simulate these “slow” biological processes.
To improve the efficiency of conformational sampling, enhanced
MD sampling techniques have been developed. Thus, these new
methodologies, combined with efficient computing hardware,
allow the study of biological processes from a structural, thermo-
dynamic and kinetic point of view in the most appropriate
biological environment [11, 12].
In this chapter, we present an overview of the main enhanced
MD methods used currently in drug design. We expose the char-
acteristics and limits of each MD method and examples of applica-
tions. We also discuss the appropriate choice of software and
hardware. We specially focus on two widely used methods, umbrella
sampling (US), a reaction coordinate (RC)-dependent method
with which the free energy profile can be computed; and acceler-
ated molecular dynamics (aMD), a nondependent RC method that
improves the conformational sampling by overcoming high energy
barriers. We give a technical, methodological, and detailed guide-
line to set up, run, and analyze US and aMD simulations. Finally,
we highlight important critical points to be analyzed for the rele-
vance and rigor of the study.2 Materials
To perform an MD simulation, we first need suitable hardware for
performing calculations, second a software package with MD cap-
abilities, and finally the initial spatial coordinates of the system of
interest. Choices of each of these three aspects are mutually depen-
dent since not all MD methodologies are implemented in all soft-
ware and not all softwares are compatible with all computer
hardware technologies. Therefore we summarize in Table1 the
software in which they are implemented and the supported tech-
nology for each MD method.Enhanced Molecular Dynamics 405