Search

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.

Spectral clustering involves placing objects into clusters based on the eigenvectors and eigenvalues of an associated matrix. The technique was first applied to molecular data by Brewer [J. Chem. Inf. Model. 2007, 47, 1727-1733] who demonstrated its use on a very small dataset of 125 COX-2 inhibitors. We have determined suitable parameters for spectral clustering using a wide variety of molecular descriptors and several datasets of a few thousand compounds and compared the results of clustering using a nonoverlapping version of Brewer's use of Sarker and Boyer's algorithm with that of Ward's and k-means clustering. We then replaced the exact eigendecomposition method with two different approximate methods and concluded that Singular Value Decomposition is the most appropriate method for clustering larger compound collections of up to 100,000 compounds. We have also used spectral clustering with the Tversky coefficient to generate two sets of clusters linked by a common set of eigenvalues and have used this novel approach to cluster sets of fragments such as those used in fragment-based drug design.

The design of new alerts, that is, collections of structural features observed to result in toxicological activity, can be a slow process and may require significant input from toxicology and chemistry experts. A method has therefore been developed to help automate alert identification by mining descriptions of activating structural features directly from toxicity data sets. The method is based on jumping emerging pattern mining which is applied to a set of toxic and nontoxic compounds that are represented using atom pair descriptors. Using the resulting jumping emerging patterns, it is possible to cluster toxic compounds into groups defined by the presence of shared structural features and to arrange the clusters into hierarchies. The methodology has been tested on a number of data sets for Ames mutagenicity, oestrogenicity, and hERG channel inhibition end points. These tests have shown the method to be effective at clustering the data sets around minimal jumping-emerging structural patterns and finding descriptions of potentially activating structural features. Furthermore, the mined structural features have been shown to be related to some of the known alerts for all three tested end points.

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.

Previous studies of the analysis of molecular matched pairs (MMPs) have often assumed that the effect of a substructural transformation on a molecular property is independent of the context (i.e., the local structural environment in which that transformation occurs). Experiments with large sets of hERG, solubility, and lipophilicity data demonstrate that the inclusion of contextual information can enhance the predictive power of MMP analyses, with significant trends (both positive and negative) being identified that are not apparent when using conventional, context-independent approaches.

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.

This paper describes a project to include explicit information about bioisosteric equivalences between pairs of fragment substructures in a system for similarity-based virtual screening. Data from the BIOSTER database show that reduced graphs provide a simple way of encoding known bioisosteric equivalences in a manner that can be used during similarity searching. Scaffold-hopping experiments with the WOMBAT database show that including such information enables similarities to be identified between the reference structures and active structures from the database that contain different, but equivalent, fragment substructures. However, such equivalences also contribute to the similarities between the reference structures and inactives, and the latter equivalences can swamp those involving the actives. This presents serious problems for the routine use of information about bioisosteric fragments in similarity-based virtual screening.

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.

Reduced graph representations of chemical structures have been shown to be effective in similarity searching applications where they offer comparable performance to other 2D descriptors in terms of recall experiments. They have also been shown to complement existing descriptors and to offer potential to scaffold hop from one chemical series to another. Various methods have been developed for quantifying the similarity between reduced graphs including fingerprint approaches, graph matching, and an edit distance method. The edit distance approach quantifies the degree of similarity of two reduced graphs based on the number and type of operations required to convert one graph to the other. An attractive feature of the edit distance method is the ability to assign different weights to different operations. For example, the mutation of an aromatic ring node to an acyclic node may be assigned a higher weight than the mutation of an aromatic ring to an aliphatic ring node. In this paper, we describe a genetic algorithm (GA) for training the weights of the different edit distance operations. The method is applied to specific activity classes extracted from the MDDR database to derive activity-class specific weights. The GA-derived weights give substantially improved results in recall experiments as compared to using weights assigned on intuition. Furthermore, such activity specific weights may provide useful structure--activity information for subsequent design efforts. In a virtual screening setting when few active compounds are known, it may be more useful to have weights that perform well across a variety of different activity classes. Thus, the GA is also trained on multiple activity classes simultaneously to derive a generalized set of weights. These more generally applicable weights also represent a substantial improvement on previous work.

This paper compares the effectiveness of similarity measures based on two-dimensional fingerprints and on molecular fields for identifying pairs of bioisosteric molecules in the BIOSTER database. The results suggest that the two types of descriptor are complementary in nature, each finding some bioisosteric pairs that are not found by the other. This conclusion is confirmed by studies of groups of BIOSTER molecules that share the same activity characteristics, and by experiments that involve combining the two types of similarity measure.

The program SELECT is presented for the design of combinatorial libraries. SELECT is based on a genetic algorithm with a multi-objective fitness function. Any number of objectives can be included, provided that they can be readily calculated. Typically, the objectives would be to maximize structural diversity while ensuring that the compounds in the library have “drug-like” properties. In the examples given, structural diversity is measured using Daylight fingerprints as descriptors and either the normalized sum of pairwise dissimilarities, calculated with the cosine coefficient, or the average nearest neighbor distance, calculated with the Tanimoto coefficient, as the measure of diversity. The objectives are specified at run time. Combinatorial libraries are selected by analyzing product space, which gives significant advantages over methods that are based on analyzing reactant space. SELECT can also be used to choose an optimal configuration for a multicomponent library. The performance of SELECT is dem...

A substructural analysis approach is used to calculate biological activity profiles, which contain weights that describe the differential occurrences of generic features (specifically, the numbers of hydrogen-bond donors and acceptors, the numbers of rotatable bonds and aromatic rings, the molecular weights, and the 2 kappa alpha descriptors) in active molecules taken from the World Drug Index and in (presumed) inactive molecules taken from the SPRESI database. Even with such simple structural descriptors, the profiles discriminate effectively between active and inactive compounds. The effectiveness of the approach is further increased by using a genetic algorithm for the calculation of the weights comprising a profile. The methods have been successfully applied to a number of different data sets.

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.

Current ring perception algorithms for use on chemical graphs concentrate on processing specific structures. In this review, the various published ring perception algorithms are classified according to the initial ring set obtained, and each algorithm or method of perception is described in detail. The final ring sets obtained are discussed in terms of their suitability for use in representing the ring systems in structurally explicit parts of generic chemical structures.

This paper reports how Zamora's smallest set of smallest rings algorithm has been modified and extended to provide an algorithm that will find the extended set of smallest rings (ESSR) for specific and structurally explicit generic structures. Modifications are necessary to find the ESSR rings within a partial structure connection table, while extensions are required to perceive the ESSR rings that span several such connection tables. Particular care is involved when structures containing primary cut faces or alternative embedments are processed

The rings perceived within and across the partial structures of structurally explicit generics are analyzed to produce ring-screen information that complements the existing fragment screens. This paper presents and discusses the format of these ring screens. The resultant bit vectors for each partial structure are accumulated to retain their logical relationships by means of a «bubble-up» of the vectors. The principles of the bubble-up are outlined and explained by means of several examples

This paper considers the problems associated with processing planar (two-dimensional) representations of three-dimensional structures. To overcome the ambiguity of such representations, a new ring set is defined in terms of simple faces and cut faces. The concept of a cut-vertex graph is introduced to explain the combinatorial relationship between the number of simple faces and the number of planar embedments