Adapting the DeepSARM approach for dual-target ligand design

The structure–activity relationship (SAR) matrix (SARM) methodology and data structure was originally developed to extract structurally related compound series from data sets of any composition, organize these series in matrices reminiscent of R-group tables, and visualize SAR patterns. The SARM approach combines the identification of structural relationships between series of active compounds with analog design, which is facilitated by systematically exploring combinations of core structures and substituents that have not been synthesized. The SARM methodology was extended through the introduction of DeepSARM, which added deep learning and generative modeling to target-based analog design by taking compound information from related targets into account to further increase structural novelty. Herein, we present the foundations of the SARM methodology and discuss how DeepSARM modeling can be adapted for the design of compounds with dual-target activity. Generating dual-target compounds represents an equally attractive and challenging task for polypharmacology-oriented drug discovery. The DeepSARM-based approach is illustrated using a computational proof-of-concept application focusing on the design of candidate inhibitors for two prominent anti-cancer targets.

     

Rationalizing the role of SAR tolerance for ligand-based virtual screening

It is well appreciated that the results of ligand-based virtual screening (LBVS) are much influenced by methodological details, given the generally strong compound class dependence of LBVS methods. It is less well understood to what extent structure-activity relationship (SAR) characteristics might influence the outcome of LBVS. We have assessed the hypothesis that the success of prospective LBVS depends on the SAR tolerance of screening targets, in addition to methodological aspects. In this context, SAR tolerance is rationalized as the ability of a target protein to specifically interact with series of structurally diverse active compounds. In compound data sets, SAR tolerance articulates itself as SAR continuity, i.e., the presence of structurally diverse compounds having similar potency. In order to analyze the role of SAR tolerance for LBVS, activity landscape representations of compounds active against 16 different target proteins were generated for which successful LBVS applications were reported. In all instances, the activity landscapes of known active compounds contained multiple regions of local SAR continuity. When analyzing the location of newly identified LBVS hits and their SAR environments, we found that these hits almost exclusively mapped to regions of distinct local SAR continuity. Taken together, these findings indicate the presence of a close link between SAR tolerance at the target level, SAR continuity at the ligand level, and the probability of LBVS success.     

Multitarget structure-activity relationships characterized by activity-difference maps and consensus similarity measure

Dual and triple activity-difference (DAD/TAD) maps are tools for the systematic characterization of structure-activity relationships (SAR) of compound data sets screened against two or three targets. DAD and TAD maps are two- and three- dimensional representations of the pairwise activity differences of compound data sets, respectively. Adding pairwise structural similarity information into these maps readily reveals activity cliff regions in the SAR for one, two, or three targets. In addition, pairs of compounds in the smooth regions of the SAR and scaffold hops are also easily identified in these maps. Herein, DAD and TAD maps are employed for the systematic characterization of the SAR of a benchmark set of 299 compounds screened against dopamine, norepinephrine, and serotonin transporters. To reduce the well-known dependence of the activity landscape on the structural representation, five selected 2D and 3D structure representations were used to characterize the SAR. Systematic analysis of the DAD and TAD maps reveals regions in the landscape with similar SAR for two or the three targets as well as regions with inverse SAR, i.e., changes in structure that increase activity for one target, but decrease activity for the other target. Focusing the analysis on pairs of compounds with high structure similarity revealed the presence of single-, dual-, and triple-target activity cliffs, i.e., small changes in structure with high changes in potency for one, two, or the three targets, respectively. Triple-target scaffold hops are also discussed. Activity cliffs and scaffold hops were also quantified and represented using two recently proposed approaches namely, mean Structure Activity Landscape Index (mean SALI) and Consensus Structure-Activity Similarity (SAS) maps.     

From activity cliffs to activity ridges: informative data structures for SAR analysis

The extraction of SAR information from structurally diverse compound data sets is a challenging task. One of the focal points of systematic SAR analysis is the search for activity cliffs, that is, structurally similar compounds having large potency differences, from which SAR determinants can be deduced. The assessment of SAR information is usually based on pairwise similarity and potency comparisons of data set compounds. As a consequence, activity cliffs are mostly evaluated at a compound pair level. Here, we present an extension of the activity cliff concept by introducing "activity ridges" that are formed by overlapping "combinatorial" activity cliffs between participating compounds, giving rise to ridge-like structures in activity landscapes. Activity ridges are rich in SAR information. In a systematic analysis of 242 compound data sets, we have identified well-defined activity ridges in 71 different sets. In addition, an information-theoretic approach has been devised to characterize the structural composition of activity ridges. Taken together, our results show that activity ridges frequently occur in sets of active compounds and that different categories of ridges can be distinguished on the basis of their structural content. The computational identification of activity ridges provides access to compound subsets having high priority for SAR analysis.     

Assessing the confidence level of public domain compound activity data and the impact of alternative potency measurements on SAR analysis

Publicly available compound activity data have been analyzed to distinguish between compounds for which single or multiple potency measurements were available and gain insight into data confidence levels. Different potency measurements with defined end points and alternative ways to represent multiple potency values for active compounds have been evaluated in the context of SAR analysis. Approximately 78% of all compounds with multiple potency measurements were found to represent high-confidence data, which corresponded to ～10% of all activity data. The use of different types of potency measurements and alternative representations of multiple potency values changed the SAR information content of compound data sets and resulted in different activity cliff distributions. Thus, the types of activity measurements that were available and how they were used substantially impacted SAR analysis. Compounds with multiple K(i) measurements provided the most reliable basis for SAR exploration.     

Target family-directed exploration of scaffolds with different SAR profiles

The scaffold concept is widely applied in chemoinformatics and medicinal chemistry to organize bioactive compounds according to common core structures or associate compound classes with specific biological activities. A variety of scaffold analyses have been carried out to derive statistics for scaffold distributions, generate structural organization schemes, or identify scaffolds that preferentially occur in given compound activity classes. Herein we further extend scaffold analysis by identifying scaffolds that display defined SAR profiles consisting of multiple properties. A structural relationship-based scaffold network has been designed as the basic data structure underlying our analysis. From network representations of scaffolds extracted from compounds active against 32 different target families, scaffolds with different SAR profiles have been extracted on the basis of decision trees that capture structural and functional characteristics of scaffolds in different ways. More than 600 scaffolds and 100 scaffold clusters were assigned to 10 SAR profiles. These scaffold sets represent different activity and target selectivity profiles and are provided for further SAR investigations including, for example, the exploration of alternative analog series for a given target of target family or the design of novel compounds on the basis of scaffold(s) with desired SAR profiles.     

Design of multitarget activity landscapes that capture hierarchical activity cliff distributions

An activity landscape model of a compound data set can be rationalized as a graphical representation that integrates molecular similarity and potency relationships. Activity landscape representations of different design are utilized to aid in the analysis of structure-activity relationships and the selection of informative compounds. Activity landscape models reported thus far focus on a single target (i.e., a single biological activity) or at most two targets, giving rise to selectivity landscapes. For compounds active against more than two targets, landscapes representing multitarget activities are difficult to conceptualize and have not yet been reported. Herein, we present a first activity landscape design that integrates compound potency relationships across multiple targets in a formally consistent manner. These multitarget activity landscapes are based on a general activity cliff classification scheme and are visualized in graph representations, where activity cliffs are represented as edges. Furthermore, the contributions of individual compounds to structure-activity relationship discontinuity across multiple targets are monitored. The methodology has been applied to derive multitarget activity landscapes for compound data sets active against different target families. The resulting landscapes identify single-, dual-, and triple-target activity cliffs and reveal the presence of hierarchical cliff distributions. From these multitarget activity landscapes, compounds forming complex activity cliffs can be readily selected.     

Compound set enrichment: a novel approach to analysis of primary HTS data

The main goal of high-throughput screening (HTS) is to identify active chemical series rather than just individual active compounds. In light of this goal, a new method (called compound set enrichment) to identify active chemical series from primary screening data is proposed. The method employs the scaffold tree compound classification in conjunction with the Kolmogorov-Smirnov statistic to assess the overall activity of a compound scaffold. The application of this method to seven PubChem data sets (containing between 9389 and 263679 molecules) is presented, and the ability of this method to identify compound classes with only weakly active compounds (potentially latent hits) is demonstrated. The analysis presented here shows how methods based on an activity cutoff can distort activity information, leading to the incorrect activity assignment of compound series. These results suggest that this method might have utility in the rational selection of active classes of compounds (and not just individual active compounds) for followup and validation.     

Activity cliffs in PubChem confirmatory bioassays taking inactive compounds into account

Activity cliffs are formed by pairs or groups of structurally similar compounds with significant differences in potency. They represent a prominent feature of activity landscapes of compound data sets and a primary source of structure–activity relationship (SAR) information. Thus far, activity cliffs have only been considered for active compounds, consistent with the principles of the activity landscape concept. However, from an SAR perspective, pairs formed by structurally similar active and inactive compounds should often also be informative. Therefore, we have extended the activity cliff concept to also take inactive compounds into consideration. As source of both confirmed active and inactive compounds, we have exclusively focused on PubChem confirmatory bioassays. Activity cliffs formed between pairs of active compounds (homogeneous pairs) and pairs of active and inactive compounds (heterogeneous pairs) were systematically analyzed on a per-assay basis, hence ensuring the currently highest possible degree of experimental consistency in activity measurement. Only very small numbers of large-magnitude activity cliffs formed between active compounds were detected in PubChem bioassays. However, when taking confirmed inactive compounds from confirmatory assays into account, the activity cliff frequency in assay data significantly increased, involving 11–15 % of all qualifying pairs of similar compounds, depending on the molecular representations that were used. Hence, these non-conventional activity cliffs provide an additional source of SAR information.     

How large does a compound screening collection need to be?

Increasingly, chemical libraries are being produced which are focused on a biological target or group of related targets, rather than simply being constructed in a combinatorial fashion. A screening collection compiled from such libraries will contain multiple analogues of a number of discrete series of compounds. The question arises as to how many analogues are necessary to represent each series in order to ensure that an active series will be identified. Based on a simple probabilistic argument and supported by in-house screening data, guidelines are given for the number of compounds necessary to achieve a "hit", or series of hits, at various levels of certainty. Obtaining more than one hit from the same series is useful since this gives early acquisition of SAR (structure-activity relationship) and confirms a hit is not a singleton. We show that screening collections composed of only small numbers of analogues of each series are sub-optimal for SAR acquisition. Based on these studies, we recommend a minimum series size of about 200 compounds. This gives a high probability of confirmatory SAR (i.e. at least two hits from the same series). More substantial early SAR (at least 5 hits from the same series) can be gained by using series of about 650 compounds each. With this level of information being generated, more accurate assessment of the likely success of the series in hit-to-lead and later stage development becomes possible.     

From bird’s eye views to molecular communities: two-layered visualization of structure–activity relationships in large compound data sets

The analysis of structure–activity relationships (SARs) becomes rather challenging when large and heterogeneous compound data sets are studied. In such cases, many different compounds and their activities need to be compared, which quickly goes beyond the capacity of subjective assessments. For a comprehensive large-scale exploration of SARs, computational analysis and visualization methods are required. Herein, we introduce a two-layered SAR visualization scheme specifically designed for increasingly large compound data sets. The approach combines a new compound pair-based variant of generative topographic mapping (GTM), a machine learning approach for nonlinear mapping, with chemical space networks (CSNs). The GTM component provides a global view of the activity landscapes of large compound data sets, in which informative local SAR environments are identified, augmented by a numerical SAR scoring scheme. Prioritized local SAR regions are then projected into CSNs that resolve these regions at the level of individual compounds and their relationships. Analysis of CSNs makes it possible to distinguish between regions having different SAR characteristics and select compound subsets that are rich in SAR information.     

The design and application of target-focused compound libraries

Target-focused compound libraries are collections of compounds which are designed to interact with an individual protein target or, frequently, a family of related targets (such as kinases, voltage-gated ion channels, serine/cysteine proteases). They are used for screening against therapeutic targets in order to find hit compounds that might be further developed into drugs. The design of such libraries generally utilizes structural information about the target or family of interest. In the absence of such structural information, a chemogenomic model that incorporates sequence and mutagenesis data to predict the properties of the binding site can be employed. A third option, usually pursued when no structural data are available, utilizes knowledge of the ligands of the target from which focused libraries can be developed via scaffold hopping. Consequently, the methods used for the design of target-focused libraries vary according to the quantity and quality of structural or ligand data that is available for each target family. This article describes examples of each of these design approaches and illustrates them with case studies, which highlight some of the issues and successes observed when screening target-focused libraries.     

Evolving interpretable structure-activity relationship models. 2. Using multiobjective optimization to derive multiple models

A multiobjective evolutionary algorithm (MOEA) is described for evolving multiple structure-activity relationships (SARs). The SARs are encoded in easy-to-interpret reduced graph queries which describe features that are preferentially present in active compounds compared to inactives. The MOEA addresses a limitation associated with many machine learning methods; that is, the inherent tradeoff that exists in recall and precision which is usually handled by combining the two objectives into a single measure with a consequent loss of control. By simultaneously optimizing recall and precision, the MOEA generates a family of SARs that lie on the precision-recall (PR) curve. The user is then able to select a query with an appropriate balance in the two objectives: for example, a low recall-high precision query may be preferred when establishing the SAR, whereas a high recall-low precision query may be more appropriate in a virtual screening context. Each query on the PR curve aims at capturing the structure-activity information into a single representation, and each can be considered as an alternative (equally valid) solution. We then investigate combining individual queries into teams with the aim of capturing multiple SARs that may exist in a data set, for example, as is commonly seen in high-throughput screening data sets. Team formation is carried out iteratively as a postprocessing step following the evolution of the individual queries. The inclusion of uniqueness as a third objective within the MOEA provides an effective way of ensuring the queries are complementary in the active compounds they describe. Substantial improvements in both recall and precision are seen for some data sets. Furthermore, the resulting queries provide more detailed structure-activity information than is present in a single query.     

Design of diverse and focused combinatorial libraries using an alternating algorithm

There is considerable research in chemistry to develop reaction conditions so that any of a very large number of reactants will successfully form new compounds, e.g. for two components, A(i) + B(j) --> A-B(ij). The numbers of A's and B's usually make it impossible to make all the possible products; with multicomponent reactions, there could easily be millions to billions of possible products. There is a need to identify subsets of reagents so that the resulting products have desirable predicted properties. Our idea is to select reactants sequentially and iteratively to optimize the evolving candidate library. The new Alternating Algorithm, AA, can be used for diversity, a space-filling design, or for a focused design, using either a near neighborhood or structure-activity relationship, SAR. A diversity design seeks to select compounds different from one another; a focused design seeks to find compounds similar to an active compound or compounds that follow a structure activity relationship. The benefit of the method is rapid computation of diversity or focused combinatorial chemical libraries.     

Activity-aware clustering of high throughput screening data and elucidation of orthogonal structure-activity relationships

From a medicinal chemistry point of view, one of the primary goals of high throughput screening (HTS) hit list assessment is the identification of chemotypes with an informative structure-activity relationship (SAR). Such chemotypes may enable optimization of the primary potency, as well as selectivity and phamacokinetic properties. A common way to prioritize them is molecular clustering of the hits. Typical clustering techniques, however, rely on a general notion of chemical similarity or standard rules of scaffold decomposition and are thus insensitive to molecular features that are enriched in biologically active compounds. This hinders SAR analysis, because compounds sharing the same pharmacophore might not end up in the same cluster and thus are not directly compared to each other by the medicinal chemist. Similarly, common chemotypes that are not related to activity may contaminate clusters, distracting from important chemical motifs. We combined molecular similarity and Bayesian models and introduce (I) a robust, activity-aware clustering approach and (II) a feature mapping method for the elucidation of distinct SAR determinants in polypharmacologic compounds. We evaluated the method on 462 dose-response assays from the Pubchem Bioassay repository. Activity-aware clustering grouped compounds sharing molecular cores that were specific for the target or pathway at hand, rather than grouping inactive scaffolds commonly found in compound series. Many of these core structures we also found in literature that discussed SARs of the respective targets. A numerical comparison of cores allowed for identification of the structural prerequisites for polypharmacology, i.e., distinct bioactive regions within a single compound, and pointed toward selectivity-conferring medchem strategies. The method presented here is generally applicable to any type of activity data and may help bridge the gap between hit list assessment and designing a medchem strategy.     

A collaborative hit-to-lead investigation leveraging medicinal chemistry expertise with high throughput library design, synthesis and purification capabilities

High throughput screening (HTS) campaigns, where laboratory automation is used to expose biological targets to large numbers of materials from corporate compound collections, have become commonplace within the lead generation phase of pharmaceutical discovery. Advances in genomics and related fields have afforded a wealth of targets such that screening facilities at larger organizations routinely execute over 100 hit-finding campaigns per year. Often, 10(5) or 10(6) molecules will be tested within a campaign/cycle to locate a large number of actives requiring follow-up investigation. Due to resource constraints at every organization, traditional chemistry methods for validating hits and developing structure activity relationships (SAR) become untenable when challenged with hundreds of hits in multiple chemical families per target. To compound the issue, comparison and prioritization of hits versus multiple screens, or physical chemical property criteria, is made more complex by the informatics issues associated with handling large data sets. This article describes a collaborative research project designed to simultaneously leverage the medicinal chemistry and drug development expertise of the Novartis Institutes for Biomedical Research Inc. (NIBRI) and ArQule Inc.'s high throughput library design, synthesis and purification capabilities. The work processes developed by the team to efficiently design, prepare, purify, assess and prioritize multiple chemical classes that were identified during high throughput screening, cheminformatics and molecular modeling activities will be detailed.     

A data mining method to facilitate SAR transfer

A challenging practical problem in medicinal chemistry is the transfer of SAR information from one chemical series to another. Currently, there are no computational methods available to rationalize or support this process. Herein, we present a data mining approach that enables the identification of alternative analog series with different core structures, corresponding substitution patterns, and comparable potency progression. Scaffolds can be exchanged between these series and new analogs suggested that incorporate preferred R-groups. The methodology can be applied to search for alternative analog series if one series is known or, alternatively, to systematically assess SAR transfer potential in compound databases.     

Systematic mining of analog series with related core structures in multi-target activity space

We have aimed to systematically extract analog series with related core structures from multi-target activity space to explore target promiscuity of closely related analogous. Therefore, a previously introduced SAR matrix structure was adapted and further extended for large-scale data mining. These matrices organize analog series with related yet distinct core structures in a consistent manner. High-confidence compound activity data yielded more than 2,300 non-redundant matrices capturing 5,821 analog series that included 4,288 series with multi-target and 735 series with multi-family activities. Many matrices captured more than three analog series with activity against more than five targets. The matrices revealed a variety of promiscuity patterns. Compound series matrices also contain virtual compounds, which provide suggestions for compound design focusing on desired activity profiles.     

Activation and Characterization of Cryptic Gene Cluster: Two Series of Aromatic Polyketides Biosynthesized by Divergent Pathways

One biosynthetic gene cluster (BGC) usually governs the biosynthesis of a series of compounds exhibiting either the same or similar molecular scaffolds. Reported here is a multiplex activation strategy to awaken a cryptic BGC associated with tetracycline polyketides, resulting in the discovery of compounds having different core structures. By constitutively expressing a positive regulator gene in tandem mode, a single BGC directed the biosynthesis of eight aromatic polyketides with two types of frameworks, two pentacyclic isomers and six glycosylated tetracyclines. The proposed biosynthetic pathway, based on systematic gene inactivation and identification of intermediates, employs two sets of tailoring enzymes with a branching point from the same intermediate. These findings not only provide new insights into the role of tailoring enzymes in the diversification of polyketides, but also highlight a reliable strategy for genome mining of natural products.     

Trend analysis of time-series data: A novel method for untargeted metabolite discovery

A new strategy for biomarker discovery is presented that uses time-series metabolomics data. Data sets from samples analysed at different time points after an intervention are searched for compounds that show a meaningful trend following the intervention. Obviously, this requires new data-analytical tools to distinguish such compounds from those showing only random variation. Two univariate methods, autocorrelation and curve-fitting, are used either as stand-alone methods or in combination to discover unknown metabolites in data sets originating from target-compound analysis. Both techniques reduce the long list of detected compounds in the kinetic sample set to include only those having a pre-defined interesting time profile. Thus, new metabolites may be discovered within data structures that are usually only used for target-compound analysis.The new strategy is tested on a sample set obtained from a gut fermentation study of a polyphenol-rich diet. For this study, the initial list of over 9000 potentially interesting features was reduced to less than 150, thus significantly reducing the expensive and time-consuming manual examination.