Introduction

Combinatorial chemistry and high-throughput screening are now important techniques in the discovery of novel bioactive compounds in the pharmaceutical and agrochemical industries. Evidence of this is seen in a recent review that lists as many as 321 combinatorial libraries that were reported in the literature in 1998 alone [1]. Combinatorial chemistry allows very large numbers of compounds to be synthesised simultaneously; however, in practice, the size of libraries is often constrained by many factors including the capacity of screening programmes and the cost of reactants. Hence, there has been a great deal of interest in developing effective methods for selecting subsets of compounds, both in the design of libraries and in compound acquisition programmes [2,3]. The criteria used for selecting compounds depend on the

* To whom correspondence should be addressed. E-mail: v,[email protected].

application for which the compounds are being used. For example, when designing libraries for screening across a wide range of biological targets the emphasis is usually on diversity so that a wide variety of structural types are contained within the library. Assessing diversity requires firstly that the compounds are described using numerical descriptors and secondly the definition of a metric that is used to quantify diversity with respect to the descriptors used [2,3]. Despite the recent flood of research into methods for assessing diversity there is still much debate about which methods are the best [4-9]. Other criteria, such as similarity to known actives, assume importance in the design of lead optimisation libraries.

In this article we describe a method we have developed for scoring and ranking compounds according to their likelihood of exhibiting activity. The method can be used to determine the order in which compounds are screened as well as to guide compound acquisition programmes. We then describe a series of experiments we have conducted that explore the benefits of designing combinatorial libraries through an analysis of product space rather than reactant space. The experiments are based on several different diversity metrics and molecular descriptors. We also show how product-based selection allows multi-objectives to be optimised simultaneously, for example, diversity and physicochemical properties, allowing the design of diverse and drug-like libraries.

0 0

Post a comment