Chemistry Reference
In-Depth Information
increasing ligand efficiency. One caveat to this is that we have not explored the magnitude
of the changes seen in these studies. A small decrease in a very ligand-efficient compound
still results in a very ligand-efficient compound.
Table 2.2
Breakdown of hit-to-lead strategies in published FBDD.
[87]
Total
Increase in LE (%) Decrease in LE (%) Unknown LE change (%)
a
23
22
61
17
Linking/assembly
NMR
9
33
67
0
X-ray
4
25
50
25
Other
10
10
60
30
Anabolic
28
46
54
0
NMR
5
60
40
0
X-ray
8
25
75
0
Other
15
53
47
0
a
Due to a lack of data, a comparison of ligand efficiencies is not possible for projects in this category.
Table 2.3
Analysis of screening methods used in published FBDD.
[87]
Screen
Total
Increase in LE (%)
Decrease in LE (%)
Unknown change in LE (%)
a
NMR
14
43
57
0
X-ray
12
25
75
0
Other
25
36
52
12
a
Due to a lack of data, a comparison of ligand efficiencies is not possible for projects in this category.
A pitfall of screening fragments by all biophysical methods, such as those already
described, is that they do not reveal whether a fragment that is found to bind to a tar-
get protein has an effect on the normal function of the protein (see below). This is why
biochemical data should be collected in parallel with biophysical data whenever possible.
2.3.2
Screen Confirmation
Many drug designers feel that functional activity must be coupled to binding of the frag-
ment. This paradigm is slowly being changed, especially as non-traditional interactions
(allosteric and/or non-active site binding sites are becoming of increasing interest and
'druggability'.
[
88, 89
]
In fact, SAR driven solely by binding information is much easier to
interpret than biochemical data because it is a direct measurement. Biochemical assays
can lead to false avenues of interest, especially with novel modes of binding/inhibition
where interpretation of the biochemical effect may be difficult. A screen should be viewed
as confirmed, not upon duplication of the original result, but when a second, orthogonal
type of data is obtained. Often it is easiest to obtain biochemical data as the orthogonal
data to biophysical data. We reiterate that biochemical data and biophysical data must be
correlated for a hit to advance out of the LG phase. After the hit has been confirmed, the
SAR hypothesis must be evaluated.
