MELLODDY: Cross-pharma Federated Learning at Unprecedented Scale Unlocks Benefits in QSAR without Compromising Proprietary Information.

Federated multipartner machine learning has been touted as an appealing and efficient method to increase the effective training data volume and thereby the predictivity of models, particularly when the generation of training data is resource-intensive. In the landmark MELLODDY project, indeed, each of ten pharmaceutical companies realized aggregated improvements on its own classification or regression models through federated learning. To this end, they leveraged a novel implementation extending multitask learning across partners, on a platform audited for privacy and security. The experiments involved an unprecedented cross-pharma data set of 2.6+ billion confidential experimental activity data points, documenting 21+ million physical small molecules and 40+ thousand assays in on-target and secondary pharmacodynamics and pharmacokinetics. Appropriate complementary metrics were developed to evaluate the predictive performance in the federated setting. In addition to predictive performance increases in labeled space, the results point toward an extended applicability domain in federated learning. Increases in collective training data volume, including by means of auxiliary data resulting from single concentration high-throughput and imaging assays, continued to boost predictive performance, albeit with a saturating return. Markedly higher improvements were observed for the pharmacokinetics and safety panel assay-based task subsets.

Discovery of New Binders for DCAF1, an Emerging Ligase Target in the Targeted Protein Degradation Field.

In this study, we describe the rapid identification of potent binders for the WD40 repeat domain (WDR) of DCAF1. This was achieved by two rounds of iterative focused screening of a small set of compounds selected on the basis of internal WDR domain knowledge followed by hit expansion. Subsequent structure-based design led to nanomolar potency binders with a clear exit vector enabling DCAF1-based bifunctional degrader exploration.

Splitting chemical structure data sets for federated privacy-preserving machine learning.

With the increase in applications of machine learning methods in drug design and related fields, the challenge of designing sound test sets becomes more and more prominent. The goal of this challenge is to have a realistic split of chemical structures (compounds) between training, validation and test set such that the performance on the test set is meaningful to infer the performance in a prospective application. This challenge is by its own very interesting and relevant, but is even more complex in a federated machine learning approach where multiple partners jointly train a model under privacy-preserving conditions where chemical structures must not be shared between the different participating parties. In this work we discuss three methods which provide a splitting of a data set and are applicable in a federated privacy-preserving setting, namely: a. locality-sensitive hashing (LSH), b. sphere exclusion clustering, c. scaffold-based binning (scaffold network). For evaluation of these splitting methods we consider the following quality criteria (compared to random splitting): bias in prediction performance, classification label and data imbalance, similarity distance between the test and training set compounds. The main findings of the paper are a. both sphere exclusion clustering and scaffold-based binning result in high quality splitting of the data sets, b. in terms of compute costs sphere exclusion clustering is very expensive in the case of federated privacy-preserving setting.

Evolution of Novartis' Small Molecule Screening Deck Design.

This article summarizes the evolution of the screening deck at the Novartis Institutes for BioMedical Research (NIBR). Historically, the screening deck was an assembly of all available compounds. In 2015, we designed a first deck to facilitate access to diverse subsets with optimized properties. We allocated the compounds as plated subsets on a 2D grid with property based ranking in one dimension and increasing structural redundancy in the other. The learnings from the 2015 screening deck were applied to the design of a next generation in 2019. We found that using traditional leadlikeness criteria (mainly MW, clogP) reduces the hit rates of attractive chemical starting points in subset screening. Consequently, the 2019 deck relies on solubility and permeability to select preferred compounds. The 2019 design also uses NIBR's experimental assay data and inferred biological activity profiles in addition to structural diversity to define redundancy across the compound sets.

Discovery of Potent, Highly Selective, and In Vivo Efficacious, Allosteric MALT1 Inhibitors by Iterative Scaffold Morphing.

MALT1 plays a central role in immune cell activation by transducing NF-κB signaling, and its proteolytic activity represents a key node for therapeutic intervention. Two cycles of scaffold morphing of a high-throughput biochemical screening hit resulted in the discovery of MLT-231, which enabled the successful pharmacological validation of MALT1 allosteric inhibition in preclinical models of humoral immune responses and B-cell lymphomas. Herein, we report the structural activity relationships (SARs) and analysis of the physicochemical properties of a pyrazolopyrimidine-derived compound series. In human T-cells and B-cell lymphoma lines, MLT-231 potently and selectively inhibits the proteolytic activity of MALT1 in NF-κB-dependent assays. Both in vitro and in vivo profiling of MLT-231 support further optimization of this in vivo tool compound toward preclinical characterization.

Systematic Chemogenetic Library Assembly.

Chemogenetic libraries, collections of well-defined chemical probes, provide tremendous value to biomedical research but require substantial effort to ensure diversity as well as quality of the contents. We have assembled a chemogenetic library by data mining and crowdsourcing institutional expertise. We are sharing our approach, lessons learned, and disclosing our current collection of 4,185 compounds with their primary annotated gene targets (https://github.com/Novartis/MoaBox). This physical collection is regularly updated and used broadly both within Novartis and in collaboration with external partners.

Gene-signature-derived IC50s/EC50s reflect the potency of causative upstream targets and downstream phenotypes.

Multiplexed gene-signature-based phenotypic assays are increasingly used for the identification and profiling of small molecule-tool compounds and drugs. Here we introduce a method (provided as R-package) for the quantification of the dose-response potency of a gene-signature as EC50 and IC50 values. Two signaling pathways were used as models to validate our methods: beta-adrenergic agonistic activity on cAMP generation (dedicated dataset generated for this study) and EGFR inhibitory effect on cancer cell viability. In both cases, potencies derived from multi-gene expression data were highly correlated with orthogonal potencies derived from cAMP and cell growth readouts, and superior to potencies derived from single individual genes. Based on our results we propose gene-signature potencies as a novel valid alternative for the quantitative prioritization, optimization and development of novel drugs.

N-aryl-piperidine-4-carboxamides as a novel class of potent inhibitors of MALT1 proteolytic activity.

Starting from a weak screening hit, potent and selective inhibitors of the MALT1 protease function were elaborated. Advanced compounds displayed high potency in biochemical and cellular assays. Compounds showed activity in a mechanistic Jurkat T cell activation assay as well as in the B-cell lymphoma line OCI-Ly3, which suggests potential use of MALT1 inhibitors in the treatment of autoimmune diseases as well as B-cell lymphomas with a dysregulated NF-κB pathway. Initially, rat pharmacokinetic properties of this compound series were dominated by very high clearance which could be linked to amide cleavage. Using a rat hepatocyte assay a good in vitro-in vivo correlation could be established which led to the identification of compounds with improved PK properties.

Identification of SPPL2a Inhibitors by Multiparametric Analysis of a High-Content Ultra-High-Throughput Screen.

The intramembrane protease signal peptide peptidase-like 2a (SPPL2a) is a potential drug target for the treatment of autoimmune diseases due to an essential role in B cells and dendritic cells. To screen a library of 1.4 million compounds for inhibitors of SPPL2a, we developed an imaging assay detecting nuclear translocation of the proteolytically released cytosolic substrate fragment. The state-of-the-art hit calling approach based on nuclear translocation resulted in numerous false-positive hits, mainly interrupting intracellular protein trafficking. To filter the false positives, we extracted 340 image-based readouts and developed a novel multiparametric analysis method that successfully triaged the primary hit list. The identified scaffolds were validated by demonstrating activity on endogenous SPPL2a and substrate CD74/p8 in B cells. The multiparametric analysis discovered diverse cellular phenotypes and provided profiles for the whole library. The principle of the presented imaging assay, the screening strategy, and multiparametric analysis are potentially applicable in future screening campaigns.

Dark chemical matter as a promising starting point for drug lead discovery.

High-throughput screening (HTS) is an integral part of early drug discovery. Herein, we focused on those small molecules in a screening collection that have never shown biological activity despite having been exhaustively tested in HTS assays. These compounds are referred to as 'dark chemical matter' (DCM). We quantified DCM, validated it in quality control experiments, described its physicochemical properties and mapped it into chemical space. Through analysis of prospective reporter-gene assay, gene expression and yeast chemogenomics experiments, we evaluated the potential of DCM to show biological activity in future screens. We demonstrated that, despite the apparent lack of activity, occasionally these compounds can result in potent hits with unique activity and clean safety profiles, which makes them valuable starting points for lead optimization efforts. Among the identified DCM hits was a new antifungal chemotype with strong activity against the pathogen Cryptococcus neoformans but little activity at targets relevant to human safety.

High-Throughput Screening Using iPSC-Derived Neuronal Progenitors to Identify Compounds Counteracting Epigenetic Gene Silencing in Fragile X Syndrome.

Fragile X syndrome (FXS) is the most common form of inherited mental retardation, and it is caused in most of cases by epigenetic silencing of the Fmr1 gene. Today, no specific therapy exists for FXS, and current treatments are only directed to improve behavioral symptoms. Neuronal progenitors derived from FXS patient induced pluripotent stem cells (iPSCs) represent a unique model to study the disease and develop assays for large-scale drug discovery screens since they conserve the Fmr1 gene silenced within the disease context. We have established a high-content imaging assay to run a large-scale phenotypic screen aimed to identify compounds that reactivate the silenced Fmr1 gene. A set of 50,000 compounds was tested, including modulators of several epigenetic targets. We describe an integrated drug discovery model comprising iPSC generation, culture scale-up, and quality control and screening with a very sensitive high-content imaging assay assisted by single-cell image analysis and multiparametric data analysis based on machine learning algorithms. The screening identified several compounds that induced a weak expression of fragile X mental retardation protein (FMRP) and thus sets the basis for further large-scale screens to find candidate drugs or targets tackling the underlying mechanism of FXS with potential for therapeutic intervention.

Glucocorticoid receptor ligands modulate Cardiovirus encephalomyocarditis virus internal ribosome entry site activity.

The use of small molecules to modulate cellular processes is a powerful approach to investigate gene function as a complement to genetic approaches. The discovery and characterization of compounds that modulate translation initiation, the rate-limiting step of protein synthesis, is important both to provide tool compounds to explore this fundamental biological process and to further evaluate protein synthesis as a therapeutic target. While most messenger ribonucleic acids (mRNAs) recruit ribosomes via their 5' cap, some viral and cellular mRNAs initiate protein synthesis via an alternative "cap-independent" mechanism utilizing internal ribosome entry sites (IRES) elements, which are complex mRNA secondary structures, localized within the 5' nontranslated region of the mRNA upstream of the AUG start codon. This report describes the design of a functional, high throughput screen of small molecules miniaturized into a 1,536-well format and performed using the luciferase reporter gene under control of the viral Cardiovirus encephalomyocarditis virus (EMCV) IRES element to identify nontoxic compounds modulating translation initiated from the EMCV IRES. One activating compound, validated in a dose response manner, has previously been shown to bind the glucocorticoid receptor (GR). Subsequent testing of additional GR modulators further supported this as the possible mechanism of action. Detailed characterization of this compound activity supported the notion that this was due to an effect at the level of translation.

Identification of cardiac glycoside molecules as inhibitors of c-Myc IRES-mediated translation.

Translation initiation is a fine-tuned process that plays a critical role in tumorigenesis. The use of small molecules that modulate mRNA translation provides tool compounds to explore the mechanism of translational initiation and to further validate protein synthesis as a potential pharmaceutical target for cancer therapeutics. This report describes the development and use of a click beetle, dual luciferase cell-based assay multiplexed with a measure of compound toxicity using resazurin to evaluate the differential effect of natural products on cap-dependent or internal ribosome entry site (IRES)-mediated translation initiation and cell viability. This screen identified a series of cardiac glycosides as inhibitors of IRES-mediated translation using, in particular, the oncogene mRNA c-Myc IRES. Treatment of c-Myc-dependent cancer cells with these compounds showed a decrease in c-Myc protein associated with a significant modulation of cell viability. These findings suggest that inhibition of IRES-mediated translation initiation may be a strategy to inhibit c-Myc-driven tumorigenesis.

Latent hit series hidden in high-throughput screening data.

Recently a novel method termed compound set enrichment (CSE) has been described that uses the activity distribution of a structural class of compounds to identify hit series from primary screening data. This report describes how this method can be used to identify such hit series, even when no hits according to conventional hit-calling methods for a given structural class are present in the data set. Such series, which were called latent hit series, were identified prospectively in a cell-based screening campaign and also in a series of retrospective analyses of publicly available data sets from PubChem. The assay used for the prospective case study was developed to identify compounds modulating protein translation directed from the internal ribosome entry site (IRES) of the encephalomyocarditis virus (EMCV) genomic RNA. The assay was designed with the ability to detect two assay readouts. The first assay readout monitors compound effects on IRES-directed translation, and the second readout monitors the cell viability and general effect on protein expression. By applying CSE separately to both of them, six validated latent hit series with apparently no effects on cell viability were identified. For each of these series, further testing of new compounds enabled identification of additional hits, also apparently with no effect on cell viability. These validated latent hit series would have been missed by a conventional cutoff-based hit-calling approach. This prospective study further supports CSE as a method for the analysis of high-throughput screening experiments.

Mining for bioactive scaffolds with scaffold networks: improved compound set enrichment from primary screening data.

Identification of meaningful chemical patterns in the increasing amounts of high-throughput-generated bioactivity data available today is an increasingly important challenge for successful drug discovery. Herein, we present the scaffold network as a novel approach for mapping and navigation of chemical and biological space. A scaffold network represents the chemical space of a library of molecules consisting of all molecular scaffolds and smaller "parent" scaffolds generated therefrom by the pruning of rings, effectively leading to a network of common scaffold substructure relationships. This algorithm provides an extension of the scaffold tree algorithm that, instead of a network, generates a tree relationship between a heuristically rule-based selected subset of parent scaffolds. The approach was evaluated for the identification of statistically significantly active scaffolds from primary screening data for which the scaffold tree approach has already been shown to be successful. Because of the exhaustive enumeration of smaller scaffolds and the full enumeration of relationships between them, about twice as many statistically significantly active scaffolds were identified compared to the scaffold-tree-based approach. We suggest visualizing scaffold networks as islands of active scaffolds.

Recent trends and observations in the design of high-quality screening collections.

The design of a high-quality screening collection is of utmost importance for the early drug-discovery process and provides, in combination with high-quality assay systems, the foundation of future discoveries. Herein, we review recent trends and observations to successfully expand the access to bioactive chemical space, including the feedback from hit assessment interviews of high-throughput screening campaigns; recent successes with chemogenomics target family approaches, the identification of new relevant target/domain families, diversity-oriented synthesis and new emerging compound classes, and non-classical approaches, such as fragment-based screening and DNA-encoded chemical libraries. The role of in silico library design approaches are emphasized.

MultiMCS: a fast algorithm for the maximum common substructure problem on multiple molecules.

Several efficient correspondence graph-based algorithms for determining the maximum common substructure (MCS) of a pair of molecules have been published in the literature. The extension of the problem to three or more molecules is however nontrivial; heuristics used to increase the efficiency in the two-molecule case are either inapplicable to the many-molecule case or do not provide significant speedups. Our specific algorithmic contribution is two-fold. First, we show how the correspondence graph approach for the two-molecule case can be generalized to obtain an algorithm that is guaranteed to find the optimum connected MCS of multiple molecules, and that runs fast on most families of molecules using a new divide-and-conquer strategy that has hitherto not been reported in this context. Second, we provide a characterization of those compound families for which the algorithm might run slowly, along with a heuristic for speeding up computations on these families. We also extend the above algorithm to a heuristic algorithm to find the disconnected MCS of multiple molecules and to an algorithm for clustering molecules into groups, with each group sharing a substantial MCS. Our methods are flexible in that they provide exquisite control on various matching criteria used to define a common substructure.

The scaffold tree: an efficient navigation in the scaffold universe.

The Scaffold Tree algorithm (J Chem Inf Model 47:47-58, 2007) allows to organize large molecular data sets by arranging sets of molecules into a unique tree hierarchy based on their scaffolds, with scaffolds forming leaf nodes of such tree. The hierarchy is created by iterative removal of rings from more complex scaffolds using chemically meaningful set of rules, until a single, root ring is obtained. The classification is deterministic, data set independent, and scales linearly with the number of compounds included in the data set. In this review we summarize the basic principles of the Scaffold Tree methodology and review its applications, which appeared in recent medicinal chemistry literature, including the use of Scaffold Trees for visualization of large chemical data sets, compound clustering, and the identification of novel bioactive molecules. References to several computer programs, including also free tools available on the Internet, allowing to perform classification and visualization of molecules based on their scaffolds are also provided.

Rule-Based Classification of Chemical Structures by Scaffold.

Databases for small organic chemical molecules usually contain millions of structures. The screening decks of pharmaceutical companies contain more than a million of structures. Nevertheless chemical substructure searching in these databases can be performed interactively in seconds. Because of this nobody has really missed structural classification of these databases for the purpose of finding data for individual chemical substructures. However, a full deck high-throughput screen produces also activity data for more than a million of substances. How can this amount of data be analyzed? Which are the active scaffolds identified by an assays? To answer such questions systematic classifications of molecules by scaffolds are needed. In this review it is described how molecules can be hierarchically classified by their scaffolds. It is explained how such classifications can be used to identify active scaffolds in an HTS data set. Once active classes are identified, they need to be visualized in the context of related scaffolds in order to understand SAR. Consequently such visualizations are another topic of this review. In addition scaffold based diversity measures are discussed and an outlook is given about the potential impact of structural classifications on a chemically aware semantic web.

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.