Biology Reference
In-Depth Information
∆
G
1
R + L
1
R L
1
∆
G
solv
∆
G
bind
R + L
2
R L
2
∆
G
2
Fig. 5.
Thermodynamic cycle.
G
bind
between
L
1
and
L
2
. In both cases, this is done by
mutating one ligand into the other, and using FEP or TI to determine
∆
(in solution)
∆
G
bind
. The method was devised in 1984,
24
and first applied to
a protein-ligand system in 1987.
25
The same approach can be used for
various applications, such as relative solvation free energies or sequence
dependence of protein-protein interactions.
34
Note that thermodynamic
cycles can be extended to multiple ligands. A related approach based on
FEP is the single-step perturbation method,
35
in which relative free ener-
gies for not-too-different compounds are estimated by perturbation from
a single simulation of an unphysical reference state that encompasses the
characteristic molecular features of the compounds.
G
solv
and
∆
4.3. Endpoint Methods
Endpoint methods, which sample only the free and bound states and
compute
G
bind
by taking a difference, have been widely used recently
to study macromolecular structural stability or association as well as
protein-ligand binding in relation with drug design (DD) applications.
These methods are attractive because of their simplicity, their low com-
putational cost compared to more exact methods such as FEP or TI, and
the fact that they can be applied to structurally diverse compounds, since
they do not need the simulation of an unphysical transformation between
molecules. However, their theoretical foundation still needs to
be strengthened, although efforts are being made in this direction.
36
∆
