Biomedical Engineering Reference
In-Depth Information
in general, different from the chemical structures of synthetic compounds [10]. In
addition, natural products may be drug candidates themselves, or they can be the
starting point of an optimization program [10,11]. Natural products are also valuable
sources to inspire the diversity-oriented synthesis (DOS) of new chemical libraries,
with the final goal to identify compounds with previously unknown biological func-
tions [12,13]. Of note, Owens and Lipinski have commented that several natural
products are bioavailable [14], and a rationale for these observations has recently
been provided [10]. In addition to commercial databases of natural products such as
the CRC Dictionary of Natural Products [15], other databases include the Traditional
Chinese Medicine (TCM) database [16] and a collection released by Specs [17]. A
unique collection of natural products can be used in computational screening through
the Drug Discovery Portal [18]. Also, combinatorial libraries combined with high
throughput and other screening methodologies continue to play a key role in drug
discovery [1,2,19]. Continuous progress in this field can be found in comprehensive
survey series in combinatorial chemistry [20].
Continued growth in the number of molecules stored in proprietary or public
databases [21] has led to the concept of
chemical space
. Comparison of the chemical
space of compound collections is important in library selection and design [22]. When
designing new libraries or screening existing libraries in approaches such as mixture-
based screening [1,19,23], conventional high-throughput [24], and structure-based or
high-content screening [25], it is relevant to consider the chemical space coverage of
the new compounds, the structural novelty, and the pharmaceutical relevance [6,26].
Systematic analysis of the chemical space of compound libraries, in particular large
collections, often requires computational approaches [27].
In this chapter we focus on chemoinformatic approaches to describing and visu-
alizing the chemical space and molecular diversity of compound databases quan-
titatively, with an emphasis on combinatorial libraries obtained from DOS, natural
products, and natural product-like data sets. In Section 10.2 we discuss the concept
of chemical space and in Section 10.3 present general aspects of chemoinformatic
methods to analyze the chemical space. In Section 10.4 we discuss applications of
chemoinformatic analysis of compound collections using a variety of approaches.
In Section 10.5 we provide examples of recent trends in computational methods to
characterize compounds libraries. In Section 10.6 we close the chapter with a general
summary and present several conclusions.
10.2 CONCEPT OF CHEMICAL SPACE
Chemical space is an intuitive concept because of its analogy with the cosmic uni-
verse. This notion is used widely in drug discovery, and it is being used increasingly
in other fields, such as flavor research [28]. However, there is not a unique definition
of chemical space. For example, in a direct relationship with the three-dimensional
physical universe, Lipinski and Hopkins state that “chemical space can be viewed
as being analogous to the cosmological universe in its vastness, with chemical com-
pounds populating space instead of stars” [29]. Using the notion of a quantitative
