Chemistry Reference
In-Depth Information
Computation tools are now fairly abundant for the design of libraries, some of which
will be described briefly below. RECAP (retrosynthetic combinatorial analysis proced-
ure) electronically fragments molecules based on chemical knowledge.
[
103
]
When applied
to databases of biologically active molecules, this allows the identification of build-
ing block fragments rich in biologically recognized elements and privileged motifs
and structures. This allows the design of building blocks and the synthesis of librar-
ies rich in biological motifs. Flux (fragment-based ligand builder reaxions) analysis
demonstrated that the fragmentation of drug-like molecules by applying simplistic pseudo-
retrosynthesis results in a stock of chemically meaningful building blocks for
de novo
molecule generation.
[
104
]
Resultant flux-designed molecules were chemically reasonable
and contained essential substructure motifs.
[
105
]
SHAPES is another method of frag-
ment analysis.
[
17, 106
]
It turns out that the acronym SHAPES does not stand for anything
but compounds were classified by shape descriptors that included their ring structure
and linker with no regard for atom type or bond order. At the basic level, SHAPES
encompasses 32 frameworks which include structures from half of all known drugs.
The program DAIM (decomposition and identification of molecules) has been developed
virtually to deconstruct compounds in small-molecule libraries for fragment-based dock-
ing and also database analysis. An advantage of using DAIM-generated fragments,
instead of filtering existing libraries for compounds with low MW, is that it keeps track
of the covalent bonds 'cleaved' in the deconstruction. This information is useful for
estimating the ease of synthesis of
de novo
-designed molecules.
[
107
]
DAIM has been suc-
cessfully used in
in silico
screening for inhibitors of -secretase and EphB4 kinase by
fragment-based high-throughput docking. Analysis of the outcome of high-throughput
screenings in addition to approved drug molecules showed recurring moieties in the
active compounds. These moieties were separated from their parent molecules and termed
'privileged fragments'.
[
108
]
Optimization of fragments.
Different terminologies have been used to identify approaches
for the elaboration and optimization processes in FBDD: SARbyNMR,
[
109
]
fragment expan-
sion, fragment linking, assembly and analogue-by-catalogue (database mining), to name
just a few.
[
52
]
These approaches can be sorted into three general categories, 'anabolic' (bond
making), 'catabolic' (molecule breaking) and 'linking' strategies (combining fragments),
and it is possible to have different hybrids of these three approaches. It should be noted
that the goal of all three approaches is a lead molecule with
in vitro
activity that maintains
a desirable LE (
>
0.4).
In the anabolic approach, fragment hits with the highest ligand efficiency are optim-
ized through deliberate medicinal chemistry. Since the medicinal chemistry is deliberate,
fragments optimized by this approach should not be encumbered by adverse properties.
As shown above, this process has the best history of maintaining the LE for a fragment.
A rule of thumb is that for every three heavy atoms added, the activity should increase by
an order of magnitude (LE
0.4).
In a catabolic or deconstructive, process, a higher-affinity inhibitor is decomposed into
fragments by retrosynthetic means. The catabolic approach is no different from typical
LLDD and does not need to be evaluated here (although, as noted below, a different
paradigm, ligand efficiency, is needed to drive the process forward). This is no differ-
ent to what is routinely done to optimize 'lead-like' molecules. Often it is assumed that
≈
