Chemistry Reference
In-Depth Information
halogens or non-second row elements (such as sulfur) do not bias the results, with false
exclusions. Although not a new idea,
[
95
]
this method has proven valuable to assess fragment
hits.
[
93
]
We prefer the simple term -log(activity)/HAC because it lends itself most easily
to 'back of an envelope' evaluations. A 100 M compound with 10 heavy atoms is more
ligand efficient than a 1 M compound with 20 heavy atoms. Starting from a compound
with only 10 heavy atoms is advantageous because it gives more available chemical space
to optimize the activity. Based upon what is known about orally available compounds, there
is a ceiling for 30-35 heavy atoms.
With the availability of structural information, rapid improvements in potency can be
rapidly obtained.
[
52, 65, 85, 86
]
SBDD is not a panacea. It is simply another method to develop
and evaluate compositional SAR. In the final analysis, compounds must be made based on
SBDD guidance that shows significant improvement from the initial hit for SBDD to have
value. Approximately 50% of all FBDD uses NMR and/or X-ray methods as the key com-
ponent
[
87
]
(Table 2.2). For that to happen, the ligand efficiency index of the fragment hits
does not have to be extremely high: molecules of 200-300 Da (14-20 HA) with potency
ranges of 0.5-20 M have a LE range of 0.23-0.45 [using then term -log(activity)/HAC].
The median ligand efficiency for all methods discussed in Tables 2.2 and 2.3 is 0.40
(
G
/HAC). This seems to indicate that good fragments (ones that advance) should start
within this range of ligand efficiency.
2.4.2
Screen Follow-up
The fragment hit follow-up process is no different to lead-like hit follow-up: the results
must be confirmed with orthogonal data, the SAR hypothesis of the screen must be eval-
uated and modified and iterative screening must commence to support further or disprove
the SAR hypothesis. To this end, additional libraries are assembled, screened and SBDD
and
in silico
efforts are engaged at a higher level.
[
7, 48, 96 100
]
As long as it is understood
that a different paradigm for evaluation of the fragments quality needs to be employed
(ligand efficiency), there is no difference between lead-like and fragment-based hit
optimization.
2.4.3 Computational Efforts/Modeling
In silico
efforts reside in Phase III activities, but can be used in every facet of the FBDD
process and comprise many computational tools.
[
101
]
SBDD approaches that have shown
promise in recent years involve the use of computational methods to find putative binding
proteins for a given compound from either genomic or protein databases and subsequently
to use experimental procedures to validate the computational result. As developed over
the past decade and summarized in a recent paper,
[
26
]
NMR-based screening of a variety
of protein targets with a large compound library demonstrates that the hot-spot regions
bind a large variety of small molecules. It was found that small compounds bind almost
exclusively to well-defined, localized regions of proteins, independent of their affinity.
Once these hot-spots have been identified, binding interactions with adjacent regions of the
protein surface can be subsequently explored to increase selectivity and improve affinity.
Druggable binding pockets generally contain small regions that are crucial to the binding
of functional groups, making them the prime targets in drug design.
[
102
]
