Chemistry Reference
In-Depth Information
Lameijer
et al
. developed a software tool to design drug-like molecules, the 'Molecule
Evoluator'.
[
44
]
In this tool, both atom- and fragment-based evolutionary approaches were
implemented. Fragments were taken from the analysis of the NCI database (ref. 27 and
reviewed in Section 8.3.1). Through
interactive
evolution, a new principle in which
the user acts as a fitness function, the authors suggested a number of simple yet novel
molecules, eight of which were subsequently synthesized. Four compounds showed affinity
for biogenic amine targets (receptor, ion channel and transport protein).
[
45
]
8.4 Conclusion
In this chapter, we have compiled a number of computational strategies to dissect molecules
into sets of constituting atoms, leading to fragments of different nature. Such fragments may
also consist of elaborate atom representations, including wildcards. The reason for doing
these, often computationally intensive, operations is found in the wealth of information
that can be gleaned from such analyses. Virtual and real-world compound libraries can be
mined for their diversity and/or similarity. In addition, the 'synthetic habits' of medicinal
chemists can be explored. Furthermore, occurrence and co-occurrence of fragments may
suggest new directions into chemical space. Fragments that appear linked to side-effects,
via either multiple activities or straight toxicity, have been identified. This may help the
medicinal chemist in designing safer or more selective lead compounds. Conversely, desired
activities can be linked to fragments and such information may be a decisive factor in a
successful medicinal chemistry program. With both the large number of HTS campaigns
being performed and the resulting data increasingly being made available in the public
domain, it is expected that steadily more dedicated datasets will become available for
fragment mining. Rule- and knowledge-based design efforts will certainly benefit from this.
References
[1] PubChem database, pubchem.ncbi.nlm.nih.gov.
[2] eMolecules, www.emolecules.com.
[3] T.I. Oprea and J.M. Blaney, Cheminformatics approaches to fragment-based lead discovery, in
Fragment-based Approaches in Drug Discovery
, ed. W. Jahnke and D.A. Erlanson, Methods
and Principles in Medicinal Chemistry, Vol. 34, Wiley-VCH Verlag GmbH, Weinheim, 2006,
pp. 91-111.
[4] R.D. Cramer, R.J. Jilek and K.M.Andrews, Dbtop: topomer similarity searching of conventional
structure databases,
J. Mol. Graph. Model
.,
20
, 447-462 (2002).
[5] P.J. Hansen and P.C. Jurs, Chemical applications of graph theory,
J. Chem. Educ
.,
65
, 574-580
(1988).
[6] R. Agrawal, T. Imielinski andA. Swami, Mining association rules between sets of items in large
databases, in
Proceedings of the International Conference on Management of Data
, ACMPress,
New York, 1993.
[7] C. Borgelt and M.R. Berthold, Mining molecular fragments: finding relevant substructures of
molecules, in
Proceedings of the International Conference on Data Mining (ICDM)
, 2002,
pp. 51-58.
[8] X. Yan and J. Han, gSpan: graph-based substructure pattern mining, in
Proceedings of the
International Conference on Data Mining (ICDM)
, Maebashi City, Japan, 2002.
