Search

Reaction-based de novo design refers to the in-silico generation of novel chemical structures by combining reagents using structural transformations derived from known reactions. The driver for using reaction-based transformations is to increase the likelihood of the designed molecules being synthetically accessible. We have previously described a reaction-based de novo design method based on reaction vectors which are transformation rules that are encoded automatically from reaction databases. A limitation of reaction vectors is that they account for structural changes that occur at the core of a reaction only, and they do not consider the presence of competing functionalities that can compromise the reaction outcome. Here, we present the development of a Reaction Class Recommender to enhance the reaction vector framework. The recommender is intended to be used as a filter on the reaction vectors that are applied during de novo design to reduce the combinatorial explosion of in-silico molecules produced while limiting the generated structures to those which are most likely to be synthesisable. The recommender has been validated using an external data set extracted from the recent medicinal chemistry literature and in two simulated de novo design experiments. Results suggest that the use of the recommender drastically reduces the number of solutions explored by the algorithm while preserving the chance of finding relevant solutions and increasing the global synthetic accessibility of the designed molecules.

Reaction-based de novo design refers to the generation of synthetically accessible molecules using transformation rules extracted from known reactions in the literature. In this context, we have previously described the extraction of reaction vectors from a reactions database and their coupling with a structure generation algorithm for the generation of novel molecules from a starting material. An issue when designing molecules from a starting material is the combinatorial explosion of possible product molecules that can be generated, especially for multistep syntheses. Here, we present the development of RENATE, a reaction-based de novo design tool, which is based on a pseudo-retrosynthetic fragmentation of a reference ligand and an inside-out approach to de novo design. The reference ligand is fragmented; each fragment is used to search for similar fragments as building blocks; the building blocks are combined into products using reaction vectors; and a synthetic route is suggested for each product molecule. The RENATE methodology is presented followed by a retrospective validation to recreate a set of approved drugs. Results show that RENATE can generate very similar or even identical structures to the corresponding input drugs, hence validating the fragmentation, search, and design heuristics implemented in the tool.

Reaction classification has often been considered an important task for many different applications, and has traditionally been accomplished using hand-coded rule-based approaches. However, the availability of large collections of reactions enables data-driven approaches to be developed. We present the development and validation of a 336-class machine learning-based classification model integrated within a Conformal Prediction (CP) framework to associate reaction class predictions with confidence estimations. We also propose a data-driven approach for "dynamic" reaction fingerprinting to maximize the effectiveness of reaction encoding, as well as developing a novel reaction classification system that organizes labels into four hierarchical levels (SHREC: Sheffield Hierarchical REaction Classification). We show that the performance of the CP augmented model can be improved by defining confidence thresholds to detect predictions that are less likely to be false. For example, the external validation of the model reports 95% of predictions as correct by filtering out less than 15% of the uncertain classifications. The application of the model is demonstrated by classifying two reaction data sets: one extracted from an industrial ELN and the other from the medicinal chemistry literature. We show how confidence estimations and class compositions across different levels of information can be used to gain immediate insights on the nature of reaction collections and hidden relationships between reaction classes.

There has been a growing interest in multitask prediction in chemoinformatics, helped by the increasing use of deep neural networks in this field. This technique is applied to multitarget data sets, where compounds have been tested against different targets, with the aim of developing models to predict a profile of biological activities for a given compound. However, multitarget data sets tend to be sparse; i.e., not all compound-target combinations have experimental values. There has been little research on the effect of missing data on the performance of multitask methods. We have used two complete data sets to simulate sparseness by removing data from the training set. Different models to remove the data were compared. These sparse sets were used to train two different multitask methods, deep neural networks and Macau, which is a Bayesian probabilistic matrix factorization technique. Results from both methods were remarkably similar and showed that the performance decrease because of missing data is at first small before accelerating after large amounts of data are removed. This work provides a first approximation to assess how much data is required to produce good performance in multitask prediction exercises. Electronic supplementary material The online version of this article (10.1186/s13321-018-0281-z) contains supplementary material, which is available to authorized users.

Emerging pattern mining techniques have been applied to datasets of Ames mutagens. The discovered patterns give rise to clusters of compounds from large and biased datasets which are used to develop new structural alerts for mutagenicity in the Derek Nexus expert system.

Molecular interaction fields provide a useful description of ligand binding propensity and have found widespread use in computer-aided drug design, for example, to characterize protein binding sites and in small molecular applications, such as three-dimensional quantitative structure-activity relationships, physicochemical property prediction, and virtual screening. However, the grids on which the field data are stored are typically very large, consisting of thousands of data points, which make them cumbersome to store and manipulate. The wavelet transform is a commonly used data compression technique, for example, in signal processing and image compression. Here we use the wavelet transform to encode molecular interaction fields as wavelet thumbnails, which represent the original grid data in significantly reduced volumes. We describe a method for aligning wavelet thumbnails based on extracting extrema from the thumbnails and subsequently use them for virtual screening. We demonstrate that wavelet thumbnails provide an effective method of capturing the three-dimensional information encoded in a molecular interaction field.

Two methods are described for biasing conformational search during pharmacophore elucidation using a multiobjective genetic algorithm (MOGA). The MOGA explores conformation on-the-fly while simultaneously aligning a set of molecules such that their pharmacophoric features are maximally overlaid. By using a clique detection method to generate overlays of precomputed conformations to initialize the population (rather than starting from random), the speed of the algorithm has been increased by 2 orders of magnitude. This increase in speed has enabled the program to be applied to greater numbers of molecules than was previously possible. Furthermore, it was found that biasing the conformations explored during search time to those found in the Cambridge Structural Database could also improve the quality of the results.

Chemical databases are routinely clustered, with the aim of grouping molecules which share similar structural features. Ideally, medicinal chemists are then able to browse a few representatives of the cluster in order to interpret the shared activity of the cluster members. However, when molecules are clustered using fingerprints, it may be difficult to decipher the structural commonalities which are present. Here, we seek to represent a cluster by means of a maximum common substructure based on the shared functionality of the cluster members. Previously, we have used reduced graphs, where each node corresponds to a generalized functional group, as topological molecular descriptors for virtual screening. In this work, we precluster a database using any clustering method. We then represent the molecules in a cluster as reduced graphs. By repeated application of a maximum common edge substructure (MCES) algorithm, we obtain one or more reduced graph cluster representatives. The sparsity of the reduced graphs means that the MCES calculations can be performed in real time. The reduced graph cluster representatives are readily interpretable in terms of functional activity and can be mapped directly back to the molecules to which they correspond, giving the chemist a rapid means of assessing potential activities contained within the cluster. Clusters of interest are then subject to a detailed R-group analysis using the same iterated MCES algorithm applied to the molecular graphs.

This paper describes the extension of our earlier multi-objective method for generating plausible pharmacophore hypotheses to incorporate partial matches. Diverse sets of molecules rarely adopt exactly the same binding mode, and so allowing the identification of partial matches allows our program to be applied to larger and more diverse datasets. The method explores the conformational space of a series of ligands simultaneously with their alignment using a multi-objective genetic algorithm (MOGA). The principles of Pareto ranking are used to evolve a diverse set of pharmacophore hypotheses that are optimised on conformational energy of the ligands, the goodness of the overlay and the volume of the overlay. A partial match is defined as a pharmacophoric feature that is present in at least two, but not all, of the ligands in the set. The number of ligands that map to a given pharmacophore point is taken into account when evaluating an overlay. The method is applied to a number of test cases extracted from the Protein Data Bank (PDB) where the true overlay is known.

Knowledge Discovery in Databases (KDD) refers to the use of methodologies from machine learning, pattern recognition, statistics, and other fields to extract knowledge from large collections of data, where the knowledge is not explicitly available as part of the database structure. In this paper, we describe four modern data mining techniques, Rough Set Theory (RST), Association Rule Mining (ARM), Emerging Pattern Mining (EP), and Formal Concept Analysis (FCA), and we have attempted to give an exhaustive list of their chemoinformatics applications. One of the main strengths of these methods is their descriptive ability. When used to derive rules, for example, in structure-activity relationships, the rules have clear physical meaning. This review has shown that there are close relationships between the methods. Often apparent differences lie in the way in which the problem under investigation has been formulated which can lead to the natural adoption of one or other method. For example, the idea of a structural alert, as a structure which is present in toxic and absent in nontoxic compounds, leads to the natural formulation of an Emerging Pattern search. Despite the similarities between the methods, each has its strengths. RST is useful for dealing with uncertain and noisy data. Its main chemoinformatics applications so far have been in feature extraction and feature reduction, the latter often as input to another data mining method, such as an Support Vector Machine (SVM). ARM has mostly been used for frequent subgraph mining. EP and FCA have both been used to mine both structural and nonstructural patterns for classification of both active and inactive molecules. Since their introduction in the 1980s and 1990s, RST, ARM, EP, and FCA have found wide-ranging applications, with many thousands of citations in Web of Science, but their adoption by the chemoinformatics community has been relatively slow. Advances, both in computer power and in algorithm development, mean that there is the potential to apply these techniques to larger data sets and thus to different problems in the future.

Pharmacophore methods provide a way of establishing a structure activity relationship for a series of known active ligands. Often, there are several plausible hypotheses that could explain the same set of ligands and, in such cases, it is important that the chemist is presented with alternatives that can be tested with different synthetic compounds. Existing pharmacophore methods involve either generating an ensemble of conformers and considering each conformer of each ligand in turn or exploring conformational space on-the-fly. The ensemble methods tend to produce a large number of hypotheses and require considerable effort to analyse the results, whereas methods that vary conformation on-the-fly typically generate a single solution that represents one possible hypothesis, even though several might exist. We describe a new method for generating multiple pharmacophore hypotheses with full conformational flexibility being explored on-the-fly. The method is based on multiobjective evolutionary algorithm techniques and is designed to search for an ensemble of diverse yet plausible overlays which can then be presented to the chemist for further investigation.

This paper reviews measures of similarity and dissimilarity between pairs of chemical molecules and the use of such measures for processing chemical databases. The applications discussed include similarity searching, database clustering and diversity analysis, focusing upon measures that are based on fragment bit-string occurrence data. The paper then discusses recent work on the calculation of similarity by aligning molecular fields and on the selection of structurally diverse subsets of chemical databases.

Predicting the risk of toxic and environmental effects of chemical compounds is of great importance to all chemical industries [1]. Expert systems have shown success in predicting toxic risk by applying established knowledge of toxicology encoded as a knowledge base of structural alerts and a reasoning model. A disadvantage of expert systems is that developing new structural alerts requires considerable time and effort from domain experts. In order to expedite this process a software tool has been developed that can automatically mine representations of activating features directly from toxicity datasets and present them in an interpretable form. Our knowledge discovery tool applies emerging pattern (EP) mining [2]: a form of association rule mining [3] that is well known to computer science, but is relatively new to chemistry [4]. The EP mining algorithm accepts any data expressed as a series of binary properties, which is divided into two classes, and extracts patterns of those properties that are frequent within the data and are more frequent in one data class compared to the other. By mining emerging patterns from toxicity datasets, encoded as fingerprints of binary descriptors, the tool generates patterns of features that distinguish toxicants from innocuous compounds. These patterns represent potentially activating features of the toxic compounds that may then be used to define new alerts. The knowledge discovery tool has been tested using a public dataset of 3489 mutagens and 2981 non-mutagens, encoded as fingerprints of approximately 2000 functional groups and ring descriptors. EPs were produced and grouped into a number of hierarchical families. Six of the EPs that represented distinct chemical classes were selected for manual inspection by a toxicology expert. Relevant literature was analysed to find a mechanistic rationale for the mined features, which resulted in four new structural alerts for in vitro mutagenicity.

A program for overlaying multiple flexible molecules has been developed. Candidate overlays are generated by a novel fingerprint algorithm, scored on three objective functions (union volume, hydrogen-bond match, and hydrophobic match), and ranked by constrained Pareto ranking. A diverse subset of the best ranked solutions is chosen using an overlay-dissimilarity metric. If necessary, the solutions can be optimised. A multi-objective genetic algorithm can be used to find additional overlays with a given mapping of chemical features but different ligand conformations. The fingerprint algorithm may also be used to produce constrained overlays, in which user-specified chemical groups are forced to be superimposed. The program has been tested on several sets of ligands, for each of which the true overlay is known from protein–ligand crystal structures. Both objective and subjective success criteria indicate that good results are obtained on the majority of these sets. Electronic supplementary material The online version of this article (doi:10.1007/s10822-012-9573-y) contains supplementary material, which is available to authorized users.

SPROUT is a computer program for constrained structure generation. It is designed to generate molecules for a range of applications in molecular recognition. The program uses a number of approximations that enable a wide variety of diverse structures to be generated. Practical use of the program is demonstrated in two examples. The first demonstrates the ability of the program to generate candidate inhibitors for a receptor site of known 3D structure, specifically the GDP binding site of p21. In the second example, structures are generated to fit a pharmacophore hypothesis that models morphine agonists.

This chapter analyses the use of similarity fusion in similarity searching of chemical databases. The ranked retrieval of molecules from a database can be modelled using both analytical and simulation approaches describing the similarities between an active reference structure and both the active and the non-active molecules in the database. The simulation model described here has the advantage that it can handle unmatched molecules, i.e., those that occur in one of the ranked similarity lists that are to be fused but that do not occur in the other. Our analyses provide insights into why the results of similarity fusion are often inconsistent when used for virtual screening.

Molecular interaction fields such as those computed by the GRID program are widely used in applications such as virtual screening, molecular docking and 3D-QSAR modelling. They characterise molecules according to their favourable interaction sites and therefore enable predictions to be made on how molecules might interact. The fields are, however, comprised of a very large number of data points which presents difficulties for many applications. For example, there are likely to be high degrees of correlation between the variables which can lead to misleading results in 3D-QSAR. We describe the use of wavelet methods for approximating such data into a much smaller number of variables. We present a number of validation experiments, including use of the approximated GRIDs in 3D-QSAR, and demonstrate that wavelet approximation at high levels of data compression preserves the information content in GRID fields while significantly reducing computational requirements.

Reduced graphs provide summary representations of chemical structures by collapsing groups of connected atoms into single nodes while preserving the topology of the original structures. This chapter reviews the extensive work that has been carried out on reduced graphs at The University of Sheffield and includes discussion of their application to the representation and search of Markush structures in patents, the varied approaches that have been implemented for similarity searching, their use in cluster representation, the different ways in which they have been applied to extract structure-activity relationships and their use in encoding bioisosteres.

Neighborhood behavior describes the extent to which small structural changes defined by a molecular descriptor are likely to lead to small property changes. This study evaluates two methods for the quantification of neighborhood behavior: the optimal diagonal method of Patterson et al. and the optimality criterion method of Horvath and Jeandenans. The methods are evaluated using twelve different types of fingerprint (both 2D and 3D) with screening data derived from several lead optimization projects at GlaxoSmithKline. The principal focus of the work is the design of chemical arrays during lead optimization, and the study hence considers not only biological activity but also important drug properties such as metabolic stability, permeability, and lipophilicity. Evidence is provided to suggest that the optimality criterion method may provide a better quantitative description of neighborhood behavior than the optimal diagonal method.

A knowledge-based approach to the de novo design of synthetically feasible molecules is described. The method is based on reaction vectors which represent the structural changes that take place at the reaction center along with the environment in which the reaction occurs. The reaction vectors are derived automatically from a database of reactions which is not restricted by size or reaction complexity. A structure generation algorithm has been developed whereby reaction vectors can be applied to previously unseen starting materials in order to suggest novel syntheses. The approach has been implemented in KNIME and is validated by reproducing known synthetic routes. We then present applications of the method in different drug design scenarios including lead optimization and library enumeration. The method offers great potential for capturing and using the growing body of data on reactions that is becoming available through electronic laboratory notebooks.

A new machine learning method is presented for extracting interpretable structure-activity relationships from screening data. The method is based on an evolutionary algorithm and reduced graphs and aims to evolve a reduced graph query (subgraph) that is present within the active compounds and absent from the inactives. The reduced graph representation enables heterogeneous compounds, such as those found in high-throughput screening data, to be captured in a single representation with the resulting query encoding structure-activity information in a form that is readily interpretable by a chemist. The application of the method is illustrated using data sets extracted from the well-known MDDR data set and GSK in-house screening data. Queries are evolved that are consistent with the known SARs, and they are also shown to be robust when applied to independent sets that were not used in training.

Three field-based similarity methods are compared in retrospective virtual screening experiments. The methods are the CatShape module of CATALYST, ROCS, and an in-house program developed at the University of Sheffield called FBSS. The programs are used in both rigid and flexible searches carried out in the MDL Drug Data Report. UNITY 2D fingerprints are also used to provide a comparison with a more traditional approach to similarity searching, and similarity based on simple whole-molecule properties is used to provide a baseline for the more sophisticated searches. Overall, UNITY 2D fingerprints and ROCS with the chemical force field option gave comparable performance and were superior to the shape-only 3D methods. When the flexible methods were compared with the rigid methods, it was generally found that the flexible methods gave slightly better results than their respective rigid methods; however, the increased performance did not justify the additional computational cost required.

The high costs associated with high-throughput screening (HTS) coupled with the limited coverage and bias of current screening collections is such that diversity analysis continues to be an important criterion in lead generation. Whereas early approaches to diversity analysis were based on traditional descriptors such as two-dimensional fingerprints a recent emphasis has been on assessing scaffold coverage to ensure that a variety of different chemotypes are represented. Moreover, whether designing diverse or focused libraries, it is widely recognised that designs should aim to achieve a balance in a number of different properties and multiobjective optimisation provides an effective way of achieving such designs.

A system is being developed, based on structure generation methods, for the design of molecules to fit a variety of constraints. The system is modular and uses techniques from artificial intelligence to solve general problems in molecule design. Heuristics are being developed based, in part, on the different kinds of constraints that might be available to the system, e.g., the shape and electrostatic nature of a cavity. The heuristics are then used to restrict the combinatorial explosion that is inherent in structure generation. The program has been tested using the appa binding site of Trypsin and a range of different structures that fit the site have been generated. Some of these structures are very closely related to appa and the predictive value of the program has also been demonstrated.

In a recent companion paper we have related the operation of simple data fusion rules used in virtual screening to a multiple integral formalism. In this paper we extend these ideas to the analysis of data fusion methods applied to real data. We examine several cases of similarity fusion using different coefficients and different representations and consider the reasons for positive or negative results in terms of the similarity distributions. Results are obtained using the SUM-, MAX- MIN-, and CombMNZ-fusion rules. We also develop a customized fusion rule, which provides an estimate of the optimal possible result for fusing multiple searches of a specific database; this shows that similarity fusion can, in principle, achieve retrieval enhancements even if this is not achieved in practice with current fusion rules. The methods are extended to analyze the comparatively successful results of group fusion with multiple actives, and we provide a rationale for the observed superiority of the MAX-rule over the SUM-rule in this context.

The large numbers of compounds that are now available in drug discovery programmes have resulted in the need for methods to select compounds, both from external suppliers and in combinatorial library design procedures. In this article we describe a method that has been 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 should be 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.

Recent developments in the technologies of HTS and combinatorial chemistry have thrown down a challenge to computational chemistry, that of maximising the chemical diversity of the compounds made and screened. This paper examines the theory behind molecular diversity analysis and includes a discussion of most of the common diversity indices, and intermolecular similarity and dissimilarity measures. The extent to which the different approaches to diversity analysis have been validated and compared is reviewed. The effects of designing diverse libraries by analysing product and reagent space are presented, and the issues surrounding the comparison of libraries and databases in diversity space are discussed.

Generation of reliable pharmacophore models is a key strategy in drug design. The quality of a pharmacophore model is known to depend on several factors, with the quality of the conformer sets used perhaps being one of the most important. The goal of this study was to compare different conformational analysis methods to determine if one was superior to the others for pharmacophore generation using Catalyst/HypoGen. The five methods selected were Catalyst/Fast, Catalyst/Best, Omega, Chem-X and MacroModel. Data sets for which Catalysts models had previously been published were selected using defined quality measures. Hypotheses were generated for each of the data sets and the performance of the different conformational analysis methods was compared using both quantitative (cost and correlation coefficients) and qualitative measures (by comparing the hypotheses in terms of the features present and their spatial relationships). Two main conclusions emerged from the study. First, it was not always possible to replicate the literature results. The reasons for these failures are explored in detail, and a template for use in publications that apply the Catalyst methodology is proposed. Second, the faster rule-based methods for conformational analysis give pharmacophore models that are just as good as, and in some cases better than, the models generated using the slower, more rigorous approaches.

This paper addresses a major issue in library design, namely how to efficiently optimize the library size (number of products) and configuration (number of reagents at each position) simultaneously with other properties such as diversity, cost, and drug-like physicochemical property profiles. These objectives are often in competition, for example, minimizing the number of reactants while simultaneously maximizing diversity, and thus present difficulties for traditional optimization methods such as genetic algorithms and simulated annealing. Here, a multiobjective genetic algorithm (MOGA) is used to vary library size and configuration simultaneously with other library properties. The result is a family of solutions that explores the tradeoffs in the objectives. This is achieved without the need to assign relative weights to the objectives. The user is then able to make an informed choice on an appropriate compromise solution. The method has been applied to two different virtual libraries: a two-component aminothiazole library and a four-component benzodiazepine library.

When designing a combinatorial library it is usually desirable to optimise multiple properties of the library simultaneously and often the properties are in competition with one another. For example, a library that is designed to be focused around a given target molecule should ideally have minimum cost and also contain molecules that are bioavailable. In this paper, we describe the program MoSELECT for multiobjective library design that is based on a multiobjective genetic algorithm (MOGA). MoSELECT searches the product-space of a virtual combinatorial library to generate a family of equivalent solutions where each solution represents a combinatorial subset of the virtual library optimised over multiple objectives. The family of solutions allows the relationships between the objectives to be explored and thus enables the library designer to make an informed choice on an appropriate compromise solution. Experiments are reported where MoSELECT has been applied to the design of various focused libraries.

Early results from screening combinatorial libraries have been disappointing with libraries either failing to deliver the improved hit rates that were expected or resulting in hits with characteristics that make them undesirable as lead compounds. Consequently, the focus in library design has shifted toward designing libraries that are optimized on multiple properties simultaneously, for example, diversity and "druglike" physicochemical properties. Here we describe the program MoSELECT that is based on a multiobjective genetic algorithm and which is able to suggest a family of solutions to multiobjective library design where all the solutions are equally valid and each represents a different compromise between the objectives. MoSELECT also allows the relationships between the different objectives to be explored with competing objectives easily identified. The library designer can then make an informed choice on which solution(s) to explore. Various performance characteristics of MoSELECT are reported based on a number of different combinatorial libraries.

Publisher Summary This chapter discusses the computational methods for the analysis of molecular diversity. Four principal types of selection method are described: cluster-based selection, partition-based selection, dissimilarity-based selection, and optimization-based selection. Cluster-based selection involves subdividing a set of molecules into groups, or clusters, that exhibit a high degree of both intra-cluster similarity and inter-cluster dissimilarity, and a diverse subset is then obtained by choosing one compound from each of the clusters in turn. Partition-based selection requires the identification of a small number of characteristics, these typically being molecular properties that would be expected to affect binding at a receptor site. The range of values for each such characteristic is sub-divided into a set of sub-ranges, and the combinatorial product of all possible sub-ranges then defines the set of cells that make up the partition. Optimization-based approaches involve defining some quantitative measure of diversity, or diversity index, and then formulating the identification of the most diverse subset as a combinatorial optimization problem.

A hierarchy of screening methods applicable to full generic structures, including generic radical terms, is described. The most general level of description is fragment screening, which includes atom and bond centred fragments, ‘bubbled-up’ from full generic structures, where the logical relationships between fragments are retained in MUST and POSS screens. Ring systems are characterised and represented as a section of the bit string in a similar logical procedure.

This paper discusses the use of networks of transputers for the matching of the labelled graphs which are used to represent chemical structures in computer-based chemical information systems; in particular, the implementation of a relaxation algorithm for chemical substructure searching is described. Tests with a doubly-linked chain of transputers suggest that near-linear speedups can be obtained by a partitioning of the database which is to be searched; lesser speedups are obtained with other network configurations. Current work is described involving the exploitation of an alternative level of parallelism in the relaxation algorithm and the parallel implementation of an algorithm for the identification of the maximal substructures common to a pair of chemical compounds.

Current approaches to the design of combinatorial libraries assume that structural diversity in the reactant pools corresponds to structural diversity in the combinatorial libraries that result from reacting these pools together. In experiments with three different published libraries, dissimilarity-based compound selection (DBCS) is applied at two levels. First, the DBCS algorithm is applied at the reactant level, a library is built, and its diversity is measured. Second, the DBCS algorithm is applied to the full set of products generated by enumeration of all the reactants and the diversity of the subset is measured. Results show that reactant-based selection, which attempts to maximize diversity in the pools, results in noticeably less diverse libraries than if the selection is performed at the product level. Experiments are reported to estimate the upperbound to diversity achievable using DBCS, and it appears that DBCS is very effective at finding maximally diverse subsets. However, applying DBCS sele...