Biomedical Engineering Reference
In-Depth Information
21.6.1.3 Develop Alerts for Toxic/Reactive Functionality in Small and
Large Molecules
Hits or leads in rare, orphan, and neglected diseases can
arise from phenotypic or mechanistic screening against commercially available
libraries. Often these screening efforts arise in an academic setting. However,
because of the disconnect between academic biology and expert medicinal
chemistry, it is essential to carry out a time-consuming medicinal chemistry
annotation of putative hits or leads before expenditure of signifi cant drug
discovery effort [18, 51]. Many companies have instituted fi lters (usually
SMARTS [SMiles ARbitrary Target Specifi cation] queries) to remove undesir-
able molecules, false positives, and frequent hitters from their HTS screening
libraries or to fi lter vendor compounds. Examples include REOS from Vertex
[54] , fi lters from GSK [55], BMS [56], and Abbott [57-59]. An academic group
in Australia has developed an extensive series of substructural features for
removal of pan assay interference compounds (PAINS) from screening librar-
ies [60]. There is as yet no coordinated or readily accessible automated method
for fi ltering compounds or alerting users to reactivity issues or for that matter
bringing the expertise of many medicinal chemists into a piece of software or
database that would identify undesirable molecules for biologists. There is
considerable need to infl uence the quality of hits and leads in public databases
and prevent experimental repetition. An analogous approach could be applied
to help optimize macromolecular properties; for example, immunogenicity can
be an important obstacle to successful protein drug therapy as antibodies to
a large-molecule drug may impact therapeutic function or pharmacokinetics
or lead to severe undesirable adverse effects in vivo [59, 60]. Various
in silico
tools and databases (e.g., Immune Epitope Database (IEDB) [61] as well as
commercial tools [62]) to identify potential for immunogenicity are available,
representing an alternative to
in vitro
or
in vivo
immunogenicity assays [63].
We could integrate the ChemAxon toolkit SMARTS-based alerts and
other rule bases to fl ag (potentially) problematic substructures within a mol-
ecule [52 - 56, 58] . We will use the state - of - the - art fi lters from references in
the previous section and dynamically be able to add additional alerts
requested from discussions with experienced medicinal chemists. We could
then develop an intuitive alerts display (Fig. 21.8). Additionally, linking to
external public databases such as PubChem and patent databases could help
provide more useful information on a hit compound that could assist in decid-
ing whether to pursue it.
21.7
DISCUSSION
We have described the development of the CDD database as a case study of
how such a tool could be used for collaborations. The tool was developed using
an agile development process which uses an integrated design-build-test
process. In the space of six years this database has become a viable technology
that has attracted many research foundations, academics, biotechs, and large-
Search WWH ::
Custom Search