Currently, the most common application of MD for hidden
pockets is the mixed solvent simulations reported in 2009 by a few
different groups [77, 78]. By including small chemical probes in
simulations of target proteins, pockets can be induced or main-
tained in an open state once discovered. Recent studies using mixed
solvents are being applied to find cryptic pockets in PPIs, build
pharmacophores for new pockets, and carry out large scale virtual
screening efforts [79–81]. In addition there are now even a few
examples of approaches combining enhanced sampling and frag-
ments to maximize both sampling of conformational selection and
the ability of probes to induce and maintain pockets [82, 83].4.3 Understanding
Conformational
Change and Searching
for Hidden
Conformations
Many proteins have functionally important conformational states
which are lightly populated and hence are often not seen in experi-
mental structural studies; alternatively some conformations are not
readily crystallized. Understanding functional motions and confor-
mations of target proteins can allow the targeting of complex
allostery or delicate selectivity for elegant therapeutic interventions
(seeNote 15).
Capture of functionally interesting conformations with
antigen-binding fragment (Fab) of antibody or single domain cam-
elid antibody (VHH) along with crystallography [84–88], NMR
[89], and electron microscopy [90] has emerged as a route to
valuable structural information. This conformational capture can
then directly link structure to function, in some cases enable the
understanding of complex allostery [91, 92]. Given this interesting
conformation capture work, discussion about the use of these Fab:
target complexes for small molecule screening against specifically
useful conformations is a logical extension [93](seeNote 16).
Further to this, biotechnology companies offering this tech-
nology have started to emerge. For example, Confo Therapeutics
focuses on generating antibody stabilized structural data for
structure-based design on GPCRs. Attempting to discover hidden
but important conformations using molecular simulation can be
very challenging but in some cases can yield crucial structural
information and thermodynamic details to complement the con-
formational capture strategy.4.3.1 Directed by
Experimental Data
It is common that we do have some experimental data about a
hidden conformation which has prompted our interest. There are
advanced examples of researchers using X-ray crystallography, X-ray
scattering (SAX), double electron-electron resonance (DEER), 2D
infra-red spectroscopy (2DIR), and nuclear magnetic resonance
(NMR) to drive conformational exploration with molecular
simulation.
Crystal structure coordinates have been used for many years as
a target to drive MD simulations. Fittingly an early approach along
these lines is called “targeted MD” [94, 95]. There are many other348 Benjamin P. Cossins et al.