Derivation and Validation of Toxicophores for Mutagenicity Prediction

Receptor-based pharmacophore modeling, virtual screening, and molecular docking studies for the discovery of novel GSK-3β inhibitors

Considering the emerging importance of glycogen synthase kinase 3 beta (GSK-3β) inhibitors in treatment of Alzheimer's disease, multi-protein structure receptor-based pharmacophore modeling was adopted to generate a 3D pharmacophore model for (GSK-3β) inhibitors. The generated 3D pharmacophore was then validated using a test set of 1235 compounds. The ZINCPharmer web tool was used to virtually screen the public ZINC database using the generated 3D pharmacophore. A set of 12,251 hits was produced and then filtered according to their lead-like properties, predicted central nervous system (CNS) activity, and Pan-assay interference compounds (PAINS) fragments to 630 compounds. Scaffold Hunter was then used to cluster the filtered compounds according to their chemical structure framework. From the different clusters, 123 compounds were selected to cover the whole chemical space of the obtained hits. The SwissADME online tool was then used to filter out the compounds with undesirable pharmacokinetic properties giving a set of 91 compounds with promising predicted pharmacodynamic and pharmacokinetic properties. To confirm their binding capability to the GSK-3β binding site, molecular docking simulations were performed for the final 91 compounds in the GSK-3β binding site. Twenty-five compounds showed acceptable binding poses that bind to the key amino acids in the binding site Asp133 and Val135 with good binding scores. The quinolin-2-one derivative ZINC67773573 was found to be a promising lead for designing new GSK-3β inhibitors for Alzheimer's disease treatment. Graphical abstract A 3D pharmacophore model for the discovery of novel (GSK-3β) inhibitors.

The Trifluoromethyl Group as a Bioisosteric Replacement of the Aliphatic Nitro Group in CB1 Receptor Positive Allosteric Modulators

The first generation of CB₁ positive allosteric modulators (e.g., ZCZ011) featured a 3-nitroalkyl-2-phenyl-indole structure. Although a small number of drugs include the nitro group, it is generally not regarded as being "drug-like", and this is particularly true for aliphatic nitro groups. There are very few case studies where an appropriate bioisostere replaced a nitro group that had a direct role in binding. This may be indicative of the difficulty of replicating its binding interactions. Herein, we report the design and synthesis of ligands targeting the allosteric binding site on the CB₁ cannabinoid receptor, in which a CF₃ group successfully replaced the aliphatic NO₂. In general, the CF₃-bearing compounds were more potent than their NO₂ equivalents and also showed improved in vitro metabolic stability. The CF₃ analogue (1) with the best balance of properties was selected for further pharmacological evaluation. Pilot in vivo studies showed that (±)-1 has similar activity to (±)-ZCZ011, with both showing promising efficacy in a mouse model of neuropathic pain.

Indole-Based Hydrazones Containing A Sulfonamide Moiety as Selective Inhibitors of Tumor-Associated Human Carbonic Anhydrase Isoforms IX and XII

Novel sulfonamidoindole-based hydrazones with a 2-(hydrazinocarbonyl)-3-phenyl-1H-indole-5-sulfonamide scaffold were synthesized and tested in enzyme inhibition assays against the tumor-associated carbonic anhydrase isoforms, hCA IX and XII, and the off-targets, hCA I and II. The compounds showed selectivity against hCA IX and XII over hCA I and II. Six compounds showed K_I values lower than 10 nM against hCA IX or XII. Molecular modeling studies were performed to suggest binding interactions between the ligand and the hCA active sites.

Dual DNA topoisomerase 1 and tyrosyl-DNA phosphodiesterase 1 inhibition for improved anticancer activity

Tyrosyl-DNA phosphodiesterase 1 (Tdp1) is a DNA repair enzyme that catalyzes the hydrolysis of the phosphodiester bond in the DNA-topoisomerase 1 (Top1) covalent complex and repairs some other 3'-end DNA adducts. Currently, Tdp1 functions as an important target in cancer drug design owing to its ability to break down various DNA adducts induced by chemotherapeutics. Tdp1 inhibitors may sensitize tumor cells to the action of Top1 poisons, thereby potentiating their effects. This mini-review summarizes findings from studies reporting the combined inhibition of Top1 and Tdp1. Two different approaches have been considered for developing such drug precursors.

Integrating bioassays, chemical analysis and in silico techniques to identify genotoxicants in surface water

Identification of hazardous compounds, as the first step of water protection and regulation, is still challenged by the difficulty to establish a linkage between toxic effects and suspected contaminants. Genotoxic compounds are one type of highly relevant toxicants in surface water, which may attack the DNA and lead to cancer in individual organism, or even damaged germ cells to be passed on to future generations. Thus, the establishment of a linkage between genotoxic effects and genotoxicant is important for environmental toxicologists and chemists. For this purpose, in the present study in silico methods were integrated with bioassays, chemical analysis and literature information to identify genotoxicants in surface water. Large volume water samples from 22 sampling sites of the Danube were collected and subjected to biological and chemical analysis. Samples from the most toxic sites (JDS32, JDS44 and JDS63) induced significant genotoxic effects in the micronucleus assay, and two of them caused mutagenicity in the Ames fluctuation assay. Chemical analysis showed that 68 chemicals were detected in these most toxic samples. Literature findings and in silico techniques using the OECD QSAR Toolbox and the ChemProp software package revealed genotoxic potentials for 29 compounds out of 68 targeted chemicals. To confirm the integrative technical data, the micronucleus assay and the Ames fluctuation assay were applied with artificial mixtures of those compounds and the raw water sample extracts. The results showed that 18 chemicals explained 48.5% of the genotoxicity in the micronucleus assay. This study highlights the capability of in silico techniques in linking adverse biological effect to suspicious hazardous compounds for the identification of toxicity drivers, and demonstrates the genotoxic potential of pollutants in the Danube.

Structure-Based Docking Studies of GLUT4 Towards Exploring Selected Phytochemicals from Solanum xanthocarpum as a Therapeutic Target for the Treatment of Cancer

BACKGROUND

In recent years, there has been an exponential increase in the global burden of cancer which has been associated with several factors including environmental influence, aging, diet, infectious agents, hormonal imbalance and chronic inflammation, among others. Cancerous cells utilize more glucose for its proliferation and survival than normal cells. Thus, the regulation of glucose consumption of cancerous cells through the inhibition of glucose transporter-4-protein (GLUT4) encoded by solute carrier family-2-member-4-gene (Slc2a4) by selected phytochemicals from Solanum xanthocarpum may serve as a new therapeutic candidate for the treatment of cancer.

METHODS

The seven identified potential inhibitors of GLUT4 from Solanum xanthocarpum were retrieved from PubChem database. Examination of their drug-likeness, toxicity prediction and molecular docking studies of these compounds with GLUT4 were carried out using online tools such as Molinspiration, PreADMET V.2.0 and Patchdock server.

RESULTS

The findings revealed that, five out of the seven compounds fulfil oral drugability of Lipinski's rule of five (RO5) while two slightly meet the criteria of RO5. Conversely, five of the compounds are predicted to be mutagen while the remaining two are predicted to be safe for the body. Additionally, stigmasterol glucoside has higher binding-affinity (7590) with GLUT4 when compared to doxorubicin (6600) the control.

CONCLUSION

These findings suggest that stigmasterol glucoside from Solanum xanthocarpum could be a promising therapeutic agent with better therapeutic efficacy than doxorubicin in the treatment of cancer via the inhibition of GLUT4.

Chemistry-Wide Association Studies (CWAS): A Novel Framework for Identifying and Interpreting Structure-Activity Relationships

Quantitative structure-activity relationships (QSAR) models are often seen as a "black box" because they are considered difficult to interpret. Meanwhile, qualitative approaches, e.g., structural alerts (SA) or read-across, provide mechanistic insight, which is preferred for regulatory purposes, but predictive accuracy of such approaches is often low. Herein, we introduce the chemistry-wide association study (CWAS) approach, a novel framework that both addresses such deficiencies and combines advantages of statistical QSAR and alert-based approaches. The CWAS framework consists of the following steps: (i) QSAR model building for an end point of interest, (ii) identification of key chemical features, (iii) determination of communities of such features disproportionately co-occurring more frequently in the active than in the inactive class, and (iv) assembling these communities to form larger (and not necessarily chemically connected) novel structural alerts with high specificity. As a proof-of-concept, we have applied CWAS to model Ames mutagenicity and Stevens-Johnson Syndrome (SJS). For the well-studied Ames mutagenicity data set, we identified 76 important individual fragments and assembled co-occurring fragments into SA both replicative of known as well as representing novel mutagenicity alerts. For the SJS data set, we identified 29 important fragments and assembled co-occurring communities into SA including both known and novel alerts. In summary, we demonstrate that CWAS provides a new framework to interpret predictive QSAR models and derive refined structural alerts for more effective design and safety assessment of drugs and drug candidates.

Meeting report, ICH M7 relevant workshop: use of (Q)SAR systems and expert judgment

Use of Quantitative Structure-Activity Relationships ((Q)SAR) prediction tools has been increasing since the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) M7 guideline was issued in June 2014. The Japanese Environmental Mutagen Society and the Bacterial Mutagenicity Study Group took the initiative of the workshop on (Q)SAR in 2016 to discuss using (Q)SAR to predict mutagenicity. The aim of the workshop was to form a common understanding on the current use of (Q)SAR tools in industry and for regulatory purposes and on the process of expert judgment. This report summarizes the general session that reviewed the use of (Q)SAR tools and the case study session that discussed expert judgment.

A large comparison of integrated SAR/QSAR models of the Ames test for mutagenicity$

Results from the Ames test are the first outcome considered to assess the possible mutagenicity of substances. Many QSAR models and structural alerts are available to predict this endpoint. From a regulatory point of view, the recommendation from international authorities is to consider the predictions of more than one model and to combine results in order to develop conclusions about the mutagenicity risk posed by chemicals. However, the results of those models are often conflicting, and the existing inconsistency in the predictions requires intelligent strategies to integrate them. In our study, we evaluated different strategies for combining results of models for Ames mutagenicity, starting from a set of 10 diverse individual models, each built on a dataset of around 6000 compounds. The novelty of our study is that we collected a much larger set of about 18,000 compounds and used the new data to build a family of integrated models. These integrations used probabilistic approaches, decision theory, machine learning, and voting strategies in the integration scheme. Results are discussed considering balanced or conservative perspectives, regarding the possible uses for different purposes, including screening of large collection of substances for prioritization.

Large-scale comparison of machine learning methods for drug target prediction on ChEMBL

Deep learning is currently the most successful machine learning technique in a wide range of application areas and has recently been applied successfully in drug discovery research to predict potential drug targets and to screen for active molecules. However, due to (1) the lack of large-scale studies, (2) the compound series bias that is characteristic of drug discovery datasets and (3) the hyperparameter selection bias that comes with the high number of potential deep learning architectures, it remains unclear whether deep learning can indeed outperform existing computational methods in drug discovery tasks. We therefore assessed the performance of several deep learning methods on a large-scale drug discovery dataset and compared the results with those of other machine learning and target prediction methods. To avoid potential biases from hyperparameter selection or compound series, we used a nested cluster-cross-validation strategy. We found (1) that deep learning methods significantly outperform all competing methods and (2) that the predictive performance of deep learning is in many cases comparable to that of tests performed in wet labs (i.e., in vitro assays).

Fragment-Based Phenotypic Lead Discovery: Cell-Based Assay to Target Leishmaniasis

A rapid and practical approach for the discovery of new chemical matter for targeting pathogens and diseases is described. Fragment-based phenotypic lead discovery (FPLD) combines aspects of traditional fragment-based lead discovery (FBLD), which involves the screening of small-molecule fragment libraries to target specific proteins, with phenotypic lead discovery (PLD), which typically involves the screening of drug-like compounds in cell-based assays. To enable FPLD, a diverse library of fragments was first designed, assembled, and curated. This library of soluble, low-molecular-weight compounds was then pooled to expedite screening. Axenic cultures of Leishmania promastigotes were screened, and single hits were then tested for leishmanicidal activity against intracellular amastigote forms in infected murine bone-marrow-derived macrophages without evidence of toxicity toward mammalian cells. These studies demonstrate that FPLD can be a rapid and effective means to discover hits that can serve as leads for further medicinal chemistry purposes or as tool compounds for identifying known or novel targets.

Reviewing Hit Discovery Literature for Difficult Targets: Glutathione Transferase Omega-1 as an Example

Early stage drug discovery reporting on relatively new or difficult targets is often associated with insufficient hit triage. Literature reviews of such targets seldom delve into the detail required to critically analyze the associated screening hits reported. Here we take the enzyme glutathione transferase omega-1 (GSTO1-1) as an example of a relatively difficult target and review the associated literature involving small-molecule inhibitors. As part of this process we deliberately pay closer-than-usual attention to assay interference and hit quality aspects. We believe this Perspective will be a useful guide for future development of GSTO1-1 inhibitors, as well serving as a template for future review formats of new or difficult targets.

Distributed Representation of Chemical Fragments

This article describes an unsupervised machine learning method for computing distributed vector representation of molecular fragments. These vectors encode fragment features in a continuous high-dimensional space and enable similarity computation between individual fragments, even for small fragments with only two heavy atoms. The method is based on a word embedding algorithm borrowed from natural language processing field, and approximately 6 million unlabeled PubChem chemicals were used for training. The resulting dense fragment vectors are in contrast to the traditional sparse "one-hot" fragment representation and capture rich relational structure in the fragment space. The vectors of small linear fragments were averaged to yield distributed vectors of bigger fragments and molecules, which were used for different tasks, e.g., clustering, ligand recall, and quantitative structure-activity relationship modeling. The distributed vectors were found to be better at clustering ring systems and recall of kinase ligands as compared to standard binary fingerprints. This work demonstrates unsupervised learning of fragment chemistry from large sets of unlabeled chemical structures and subsequent application to supervised training on relatively small data sets of labeled chemicals.

In Silico Prediction of Chemical Toxicity for Drug Design Using Machine Learning Methods and Structural Alerts

During drug development, safety is always the most important issue, including a variety of toxicities and adverse drug effects, which should be evaluated in preclinical and clinical trial phases. This review article at first simply introduced the computational methods used in prediction of chemical toxicity for drug design, including machine learning methods and structural alerts. Machine learning methods have been widely applied in qualitative classification and quantitative regression studies, while structural alerts can be regarded as a complementary tool for lead optimization. The emphasis of this article was put on the recent progress of predictive models built for various toxicities. Available databases and web servers were also provided. Though the methods and models are very helpful for drug design, there are still some challenges and limitations to be improved for drug safety assessment in the future.

Identification of mutagenic transformation products generated during oxidation of 3-methyl-4-nitrophenol solutions by orbitrap tandem mass spectrometry and quantitative structure-activity relationship analyses

We used Ames assays to investigate the effects of ozonation (designated O₃), ozonation followed by chlorination (O₃/Cl), an advanced oxidation process (AOP, UV/H₂O₂), and AOP followed by chlorination (AOP/Cl) on the mutagenicity of solutions of 3-methyl-4-nitrophenol (3M4NP), a major environmental degradation product of the organophosphorus insecticide fenitrothion. Whereas O₃ did not induce mutagenicity, O₃/Cl, AOP, and AOP/Cl converted 3M4NP into mutagenic transformation products (TPs). Using liquid chromatography-mass spectrometry, we detected a total of 138 peaks in the solutions subjected to O₃/Cl, AOP, and AOP/Cl. To elucidate the TPs responsible for the observed mutagenicity, we performed simple regression analyses of the relationship between the area of each peak and the observed mutagenicity of samples withdrawn periodically during each oxidation process. The area of each of 10 peaks was found to be positively correlated (r² ≥ 0.8) with the observed mutagenicity, suggesting that the TPs corresponding to these peaks contributed to the mutagenicity. After taking into account the consistency of mutagenicity induction by the oxidation processes and analyzing the peaks by tandem mass spectrometry, we identified 3 TPs, corresponding to 6 peaks, as candidate mutagens. These TPs were assessed by means of 4 quantitative structure-activity relationship (QSAR) models, and all 3 were predicted to be mutagenic by at least one model. This result was consistent with our assumption that these TPs were mutagens. Ames assays of an authentic sample of one of the 3 TPs revealed that it did not contribute to the mutagenicity. This left 3-methoxy-4-nitrophenol and 2-[(E)-[(2,5-dihydroxyphenyl) methylidene]amino]-5-dihydroxybenzaldehyde on the list of mutagens suspected of contributing to the mutagenicity induced by AOP. No TPs were identified as candidate mutagens responsible for the mutagenicity induced by O₃/Cl and AOP/Cl.

Combining machine learning models of in vitro and in vivo bioassays improves rat carcinogenicity prediction

In vitro genotoxicity bioassays are cost-efficient methods of assessing potential carcinogens. However, many genotoxicity bioassays are inappropriate for detecting chemicals eliciting non-genotoxic mechanisms, such as tumour promotion, this necessitates the use of in vivo rodent carcinogenicity (IVRC) assays. In silico IVRC modelling could potentially address the low throughput and high cost of this assay. We aimed to develop and combine computational QSAR models of novel bioassays for the prediction of IVRC results and compare with existing software. QSAR models were generated from existing Ames (n = 6512), Syrian Hamster Embryonic (SHE, n = 410), ISSCAN rodent carcinogenicity (ISC, n = 834) and GreenScreen GADD45a-GFP (n = 1415) chemical datasets. These models mapped the molecular descriptors of each compound to their respective assay result using machine learning algorithms (adaboost, k-Nearest Neighbours, C.45 Decision Tree, Multilayer Perceptron, Random Forest). The best performing models were combined with k-Nearest Neighbours to create a cascade model for IVRC prediction. High QSAR model performance was observed from ten time 10-fold cross-validation with above 80% accuracy and 0.85 AUC for each assay dataset. The cascade model predicted rat carcinogenicity with 69.3% accuracy and 0.700 AUC. This study demonstrates the novelty of a combined approach for IVRC prediction, with higher performance than existing software.

Tumor site concordance and genetic toxicology test correlations in NTP 2-year gavage, drinking water, dermal, and intraperitoneal injection studies

The National Toxicology Program has conducted 594, 2-year studies exposing various strains of rats and mice via different routes of exposure. In the current study, we analyze the results from 108 chemicals tested in 106, 2-year studies conducted by exposing F334/N rats and B6C3F1 mice via gavage. An additional 18, 2-year gavage studies have been conducted in Osborne–Mendel rats and B6C3F1 mice on 19 different chemicals. We analyze the results from 23 chemicals tested in 21, 2-year studies conducted by exposing F334/N rats and B6C3F1 mice via drinking water; 18 chemicals tested in 18, 2-year studies conducted by exposing F334/N rats and B6C3F1 mice via dermal application; and 11 chemicals tested in 11, 2-year studies conducted by exposing F334/N rats and B6C3F1 mice via intraperitoneal injection. The results from these 174 studies are analyzed and discussed separately. The neoplasticity of each chemical was analyzed for tumor incidence by species–sex category, tumor site concordance across species, and tumor site concordance across sex within species. When available the Ames Salmonella mutagenicity assay results, and any results from a test for genotoxicity other than the Ames test, were correlated with the neoplasticity results. Tumor site concordance across sex within species is generally higher than tumor site concordance across species. In addition, the high degree of variability of Ames test results suggests that historical Ames test data are less reliable than recent results conducted under good laboratory practices and employing Organization for Economic Cooperation and Development protocols relevant to the physicochemical characteristics of the test chemical.

Applications of Chemoinformatics in Predictive Toxicology for Regulatory Purposes, Especially in the Context of the EU REACH Legislation

Chemoinformatics methodologies such as QSAR/QSPR have been used for decades in drug discovery projects, especially for the finding of new compounds with therapeutic properties and the optimization of ADME properties on chemical series. The application of computational techniques in predictive toxicology is much more recent, and they are experiencing an increasingly interest because of the new legal requirements imposed by national and international regulations. In the pharmaceutical field, the US Food and Drug Administration (FDA) support the use of predictive models for regulatory decision-making when assessing the genotoxic and carcinogenic potential of drug impurities. In Europe, the REACH legislation promotes the use of QSAR in order to reduce the huge amount of animal testing needed to demonstrate the safety of new chemical entities subjected to registration, provided they meet specific conditions to ensure their quality and predictive power. In this review, the authors summarize the state of art of in silico methods for regulatory purposes, with especial emphasis on QSAR models.

Computational Toxicology and Drug Discovery

The use of computational toxicology methods within drug discovery began in the early 2000s with applications such as predicting bacterial mutagenicity and hERG inhibition. The field has been continuously expanding ever since and the tasks at hand have become more complex. These approaches are now strategically integrated into the risk assessment process, as a complement to in vitro and in vivo methods. Today, computational toxicology can be used in every phase of drug discovery and development, from profiling large libraries early on, to predicting off-target effects in the mid-discovery phase, to assessing potential mutagenic impurities in development and degradants as part of life-cycle management. This chapter provides an overview of the field and describes the application of computational toxicology throughout the entire discovery and development process.

Computational Toxicology Methods in Chemical Library Design and High-Throughput Screening Hit Validation

The discovery of molecular toxicity in a clinical drug candidate can have a significant impact on both the cost and timeline of the drug discovery process. Early identification of potentially toxic compounds during screening library preparation or, alternatively, during the hit validation process, is critical to ensure that valuable time and resources are not spent pursuing compounds that may possess a high propensity for human toxicity. This chapter focuses on the application of computational molecular filters, applied either prescreening or postscreening, to identify and remove known reactive and/or potentially toxic compounds from consideration in drug discovery campaigns.

The Use of In Silico Models Within a Large Pharmaceutical Company

The present contribution describes how in silico models are applied at different stages of the drug discovery process in the pharmaceutical industry. A thorough description of the most relevant computational methods and tools is given along with an in-depth evaluation of their performance in the context of potential genotoxic impurities assessment.The challenges of predicting the outcome of highly complex studies are discussed followed by considerations on how novel ways to manage, store, share and analyze data may advance knowledge and facilitate modeling efforts.

Use of Read-Across Tools

Read-across has become popular since the introduction of regulations, such as the European REACH regulation. This chapter provides instructions on how to use ToxRead, new freely available software for read-across analysis, and on how to interpret its output predictions for mutagenicity assessments.This tool offers two seminal sources: a set of rules/structural alerts, which may explain the toxicity, and a similarity tool, associated with a large database of chemicals with their properties.

Discovery of a Potent, Selective T-type Calcium Channel Blocker as a Drug Candidate for the Treatment of Generalized Epilepsies

We report here the discovery and pharmacological characterization of N-(1-benzyl-1H-pyrazol-3-yl)-2-phenylacetamide derivatives as potent, selective, brain-penetrating T-type calcium channel blockers. Optimization focused mainly on solubility, brain penetration, and the search for an aminopyrazole metabolite that would be negative in an Ames test. This resulted in the preparation and complete characterization of compound 66b (ACT-709478), which has been selected as a clinical candidate.

Marine Inspired 2-(5-Halo-1H-indol-3-yl)-N,N-dimethylethanamines as Modulators of Serotonin Receptors: An Example Illustrating the Power of Bromine as Part of the Uniquely Marine Chemical Space

In previous studies, we have isolated several marine indole alkaloids and evaluated them in the forced swim test (FST) and locomotor activity test, revealing their potential as antidepressant and sedative drug leads. Amongst the reported metabolites to display such activities was 5-bromo-N,N-dimethyltryptamine. Owing to the importance of the judicious introduction of halogens into drug candidates, we synthesized two series built on a 2-(1H-indol-3-yl)-N,N-dimethylethanamine scaffold with different halogen substitutions. The synthesized compounds were evaluated for their in vitro and in vivo antidepressant and sedative activities using the mouse forced swim and locomotor activity tests. Receptor binding studies of these compounds to serotonin (5-HT) receptors were conducted. Amongst the prepared compounds, 2-(1H-indol-3-yl)-N,N-dimethyl-2-oxoacetamide (1a), 2-(5-bromo-1H-indol-3-yl)-N,N-dimethyl-2-oxoacetamide (1d), 2-(1H-indol-3-yl)-N,N-dimethylethanamine (2a), 2-(5-chloro-1H-indol-3-yl)-N,N-dimethylethanamine (2c), 2-(5-bromo-1H-indol-3-yl)-N,N-dimethylethanamine (2d), and 2-(5-iodo-1H-indol-3-yl)-N,N-dimethylethanamine (2e) have been shown to possess significant antidepressant-like action, while compounds 2c, 2d, and 2e exhibited potent sedative activity. Compounds 2a, 2c, 2d, and 2e showed nanomolar affinities to serotonin receptors 5-HT_1A and 5-HT₇. The in vitro data indicates that the antidepressant action exerted by these compounds in vivo is mediated, at least in part, via interaction with serotonin receptors. The data presented here shows the valuable role that bromine plays in providing novel chemical space and electrostatic interactions. Bromine is ubiquitous in the marine environment and a common element of marine natural products.

Comparison of the Predictive Performance and Interpretability of Random Forest and Linear Models on Benchmark Data Sets

The ability to interpret the predictions made by quantitative structure-activity relationships (QSARs) offers a number of advantages. While QSARs built using nonlinear modeling approaches, such as the popular Random Forest algorithm, might sometimes be more predictive than those built using linear modeling approaches, their predictions have been perceived as difficult to interpret. However, a growing number of approaches have been proposed for interpreting nonlinear QSAR models in general and Random Forest in particular. In the current work, we compare the performance of Random Forest to those of two widely used linear modeling approaches: linear Support Vector Machines (SVMs) (or Support Vector Regression (SVR)) and partial least-squares (PLS). We compare their performance in terms of their predictivity as well as the chemical interpretability of the predictions using novel scoring schemes for assessing heat map images of substructural contributions. We critically assess different approaches for interpreting Random Forest models as well as for obtaining predictions from the forest. We assess the models on a large number of widely employed public-domain benchmark data sets corresponding to regression and binary classification problems of relevance to hit identification and toxicology. We conclude that Random Forest typically yields comparable or possibly better predictive performance than the linear modeling approaches and that its predictions may also be interpreted in a chemically and biologically meaningful way. In contrast to earlier work looking at interpretation of nonlinear QSAR models, we directly compare two methodologically distinct approaches for interpreting Random Forest models. The approaches for interpreting Random Forest assessed in our article were implemented using open-source programs that we have made available to the community. These programs are the rfFC package ( https://r-forge.r-project.org/R/?group_id=1725 ) for the R statistical programming language and the Python program HeatMapWrapper [ https://doi.org/10.5281/zenodo.495163 ] for heat map generation.

Navigating through the minefield of read-across tools: A review of in silico tools for grouping

Read-across is a popular data gap filling technique used within analogue and category approaches for regulatory purposes. In recent years there have been many efforts focused on the challenges involved in read-across development, its scientific justification and documentation. Tools have also been developed to facilitate read-across development and application. Here, we describe a number of publicly available read-across tools in the context of the category/analogue workflow and review their respective capabilities, strengths and weaknesses. No single tool addresses all aspects of the workflow. We highlight how the different tools complement each other and some of the opportunities for their further development to address the continued evolution of read-across.

Design and Synthesis of Chlorinated and Fluorinated 7-Azaindenoisoquinolines as Potent Cytotoxic Anticancer Agents That Inhibit Topoisomerase I

The 7-azaindenoisoquinolines are cytotoxic topoisomerase I (Top1) inhibitors. Previously reported representatives bear a 3-nitro group. The present report documents the replacement of the potentially genotoxic 3-nitro group by 3-chloro and 3-fluoro substituents, resulting in compounds with high Top1 inhibitory activities and potent cytotoxicities in human cancer cell cultures and reduced lethality in an animal model. Some of the new Top1 inhibitors also possess moderate inhibitory activities against tyrosyl-DNA phosphodiesterase 1 (TDP1) and tyrosyl-DNA phosphodiesterase 2 (TDP2), two enzymes that are involved in DNA damage repair resulting from Top1 inhibitors, and they produce significantly more DNA damage in cancer cells than in normal cells. Eighteen of the new compounds had cytotoxicity mean-graph midpoint (MGM) GI50 values in the submicromolar (0.033-0.630 μM) range. Compounds 16b and 17b are the most potent in human cancer cell cultures with MGM GI50 values of 0.063 and 0.033 μM, respectively. Possible binding modes to Top1 and TDP1were investigated by molecular modeling.