Chemical Similarity Searching

Ligand-Based Virtual Screening as a Path to New Chemotypes for Candidate PET Radioligands for Imaging Tauopathies

Ligand-based virtual screening (LBVS) has rarely been tested as a method for discovering new structural scaffolds for PET radioligand development. This study used LBVS to discover potential chemotype leads for developing radioligands for PET imaging of tauopathies. ZINC12, a free database of over 12 million commercially available compounds, was searched to discover novel scaffolds based on similarities to four query compounds. Thirteen high-ranking hits were purchased and assayed for their ability to compete against three tritiated radioligands at their distinct binding sites in Alzheimer's disease brain tissue. Three hits were 2-substituted 6-methoxy naphthalenes. Synthetic elaboration of this new chemotype yielded three new ligands (25, 26, and 28) with high affinity for the [3H]6 (flortaucipur) neurofibrillary tangle binding site. Compound 28 showed remarkably high affinity (Ki, 7 nM) and other desirable properties for a candidate PET radioligand, including low topological polar surface area, moderate computed log D, and amenability for labeling with carbon-11. LBVS appears to be uniquely valuable for discovering new chemotypes for candidate PET radioligands.

Unlocking the Potential of Clustering and Classification Approaches: Navigating Supervised and Unsupervised Chemical Similarity

BACKGROUND

The field of toxicology has witnessed substantial advancements in recent years, particularly with the adoption of new approach methodologies (NAMs) to understand and predict chemical toxicity. Class-based methods such as clustering and classification are key to NAMs development and application, aiding the understanding of hazard and risk concerns associated with groups of chemicals without additional laboratory work. Advances in computational chemistry, data generation and availability, and machine learning algorithms represent important opportunities for continued improvement of these techniques to optimize their utility for specific regulatory and research purposes. However, due to their intricacy, deep understanding and careful selection are imperative to align the adequate methods with their intended applications.

OBJECTIVES

This commentary aims to deepen the understanding of class-based approaches by elucidating the pivotal role of chemical similarity (structural and biological) in clustering and classification approaches (CCAs). It addresses the dichotomy between general end point-agnostic similarity, often entailing unsupervised analysis, and end point-specific similarity necessitating supervised learning. The goal is to highlight the nuances of these approaches, their applications, and common misuses.

DISCUSSION

Understanding similarity is pivotal in toxicological research involving CCAs. The effectiveness of these approaches depends on the right definition and measure of similarity, which varies based on context and objectives of the study. This choice is influenced by how chemical structures are represented and the respective labels indicating biological activity, if applicable. The distinction between unsupervised clustering and supervised classification methods is vital, requiring the use of end point-agnostic vs. end point-specific similarity definition. Separate use or combination of these methods requires careful consideration to prevent bias and ensure relevance for the goal of the study. Unsupervised methods use end point-agnostic similarity measures to uncover general structural patterns and relationships, aiding hypothesis generation and facilitating exploration of datasets without the need for predefined labels or explicit guidance. Conversely, supervised techniques demand end point-specific similarity to group chemicals into predefined classes or to train classification models, allowing accurate predictions for new chemicals. Misuse can arise when unsupervised methods are applied to end point-specific contexts, like analog selection in read-across, leading to erroneous conclusions. This commentary provides insights into the significance of similarity and its role in supervised classification and unsupervised clustering approaches. https://doi.org/10.1289/EHP14001.

A SAR and QSAR study on 3CLpro inhibitors of SARS-CoV-2 using machine learning methods

The 3C-like Proteinase (3CLpro) of novel coronaviruses is intricately linked to viral replication, making it a crucial target for antiviral agents. In this study, we employed two fingerprint descriptors (ECFP_4 and MACCS) to comprehensively characterize 889 compounds in our dataset. We constructed 24 classification models using machine learning algorithms, including Support Vector Machine (SVM), Random Forest (RF), extreme Gradient Boosting (XGBoost), and Deep Neural Networks (DNN). Among these models, the DNN- and ECFP_4-based Model 1D_2 achieved the most promising results, with a remarkable Matthews correlation coefficient (MCC) value of 0.796 in the 5-fold cross-validation and 0.722 on the test set. The application domains of the models were analysed using dSTD-PRO calculations. The collected 889 compounds were clustered by K-means algorithm, and the relationships between structural fragments and inhibitory activities of the highly active compounds were analysed for the 10 obtained subsets. In addition, based on 464 3CLpro inhibitors, 27 QSAR models were constructed using three machine learning algorithms with a minimum root mean square error (RMSE) of 0.509 on the test set. The applicability domains of the models and the structure-activity relationships responded from the descriptors were also analysed.

Hilbert-curve assisted structure embedding method

MOTIVATION

Chemical space embedding methods are widely utilized in various research settings for dimensional reduction, clustering and effective visualization. The maps generated by the embedding process can provide valuable insight to medicinal chemists in terms of the relationships between structural, physicochemical and biological properties of compounds. However, these maps are known to be difficult to interpret, and the ''landscape'' on the map is prone to ''rearrangement'' when embedding different sets of compounds.

RESULTS

In this study we present the Hilbert-Curve Assisted Space Embedding (HCASE) method which was designed to create maps by organizing structures according to a logic familiar to medicinal chemists. First, a chemical space is created with the help of a set of ''reference scaffolds''. These scaffolds are sorted according to the medicinal chemistry inspired Scaffold-Key algorithm found in prior art. Next, the ordered scaffolds are mapped to a line which is folded into a higher dimensional (here: 2D) space. The intricately folded line is referred to as a pseudo-Hilbert-Curve. The embedding of a compound happens by locating its most similar reference scaffold in the pseudo-Hilbert-Curve and assuming the respective position. Through a series of experiments, we demonstrate the properties of the maps generated by the HCASE method. Subjects of embeddings were compounds of the DrugBank and CANVASS libraries, and the chemical spaces were defined by scaffolds extracted from the ChEMBL database.

SCIENTIFIC CONTRIBUTION

The novelty of HCASE method lies in generating robust and intuitive chemical space embeddings that are reflective of a medicinal chemist's reasoning, and the precedential use of space filling (Hilbert) curve in the process.

AVAILABILITY

https://github.com/ncats/hcase.

Stereoisomers Are Not Machine Learning's Best Friends

This study addresses the challenge of accurately identifying stereoisomers in cheminformatics, which originates from our objective to apply machine learning to predict the association constant between cyclodextrin and a guest. Identifying stereoisomers is indeed crucial for machine learning applications. Current tools offer various molecular descriptors, including their textual representation as Isomeric SMILES that can distinguish stereoisomers. However, such representation is text-based and does not have a fixed size, so a conversion is needed to make it usable to machine learning approaches. Word embedding techniques can be used to solve this problem. Mol2vec, a word embedding approach for molecules, offers such a conversion. Unfortunately, it cannot distinguish between stereoisomers due to its inability to capture the spatial configuration of molecular structures. This study proposes several approaches that use word embedding techniques to handle molecular discrimination using stereochemical information on molecules or considering Isomeric SMILES notation as a text in Natural Language Processing. Our aim is to generate a distinct vector for each unique molecule, correctly identifying stereoisomer information in cheminformatics. The proposed approaches are then compared to our original machine learning task: predicting the association constant between cyclodextrin and a guest molecule.

Reaction rebalancing: a novel approach to curating reaction databases

PURPOSE

Reaction databases are a key resource for a wide variety of applications in computational chemistry and biochemistry, including Computer-aided Synthesis Planning (CASP) and the large-scale analysis of metabolic networks. The full potential of these resources can only be realized if datasets are accurate and complete. Missing co-reactants and co-products, i.e., unbalanced reactions, however, are the rule rather than the exception. The curation and correction of such incomplete entries is thus an urgent need.

METHODS

The SynRBL framework addresses this issue with a dual-strategy: a rule-based method for non-carbon compounds, using atomic symbols and counts for prediction, alongside a Maximum Common Subgraph (MCS)-based technique for carbon compounds, aimed at aligning reactants and products to infer missing entities.

RESULTS

The rule-based method exceeded 99% accuracy, while MCS-based accuracy varied from 81.19 to 99.33%, depending on reaction properties. Furthermore, an applicability domain and a machine learning scoring function were devised to quantify prediction confidence. The overall efficacy of this framework was delineated through its success rate and accuracy metrics, which spanned from 89.83 to 99.75% and 90.85 to 99.05%, respectively.

CONCLUSION

The SynRBL framework offers a novel solution for recalibrating chemical reactions, significantly enhancing reaction completeness. With rigorous validation, it achieved groundbreaking accuracy in reaction rebalancing. This sets the stage for future improvement in particular of atom-atom mapping techniques as well as of downstream tasks such as automated synthesis planning.

SCIENTIFIC CONTRIBUTION

SynRBL features a novel computational approach to correcting unbalanced entries in chemical reaction databases. By combining heuristic rules for inferring non-carbon compounds and common subgraph searches to address carbon unbalance, SynRBL successfully addresses most instances of this problem, which affects the majority of data in most large-scale resources. Compared to alternative solutions, SynRBL achieves a dramatic increase in both success rate and accurary, and provides the first freely available open source solution for this problem.

The Chameleon Strategy-A Recipe for Effective Ligand Screening for Viral Targets Based on Four Novel Structure-Binding Strength Indices

The RNA viruses SARS-CoV, SARS-CoV-2 and MERS-CoV encode the non-structural Nsp16 (2'-O-methyltransferase) that catalyzes the transfer of a methyl group from S-adenosylmethionine (SAM) to the first ribonucleotide in mRNA. Recently, it has been found that breaking the bond between Nsp16 and SAM substrate results in the cessation of mRNA virus replication. To date, only a limited number of such inhibitors have been identified, which can be attributed to a lack of an effective "recipe". The aim of our study was to propose and verify a rapid and effective screening protocol dedicated to such purposes. We proposed four new indices describing structure-binding strength (structure-binding affinity, structure-hydrogen bonding, structure-steric and structure-protein-ligand indices) were then applied and shown to be extremely helpful in determining the degree of increase or decrease in binding affinity in response to a relatively small change in the ligand structure. After initial pre-selection, based on similarity to SAM, we limited the study to 967 compounds, so-called molecular chameleons. They were then docked in the Nsp16 protein pocket, and 10 candidate ligands were selected using the novel structure-binding affinity index. Subsequently the selected 10 candidate ligands and 8 known inhibitors and were docked to Nsp16 pockets from SARS-CoV-2, MERS-CoV and SARS-CoV. Based on the four new indices, the best ligands were selected and a new one was designed by tuning them. Finally, ADMET profiling and molecular dynamics simulations were performed for the best ligands. The new structure-binding strength indices can be successfully applied not only to screen and tune ligands, but also to determine the effectiveness of the ligand in response to changes in the target viral entity, which is particularly useful for assessing drug effectiveness in the case of alterations in viral proteins. The developed approach, the so-called chameleon strategy, has the capacity to introduce a novel universal paradigm to the field of drugs design, including RNA antivirals.

De Novo Design of Inhibitors of DNA Methyltransferase 1: A Critical Comparison of Ligand- and Structure-Based Approaches

Designing and developing inhibitors against the epigenetic target DNA methyltransferase (DNMT) is an attractive strategy in epigenetic drug discovery. DNMT1 is one of the epigenetic enzymes with significant clinical relevance. Structure-based de novo design is a drug discovery strategy that was used in combination with similarity searching to identify a novel DNMT inhibitor with a novel chemical scaffold and warrants further exploration. This study aimed to continue exploring the potential of de novo design to build epigenetic-focused libraries targeted toward DNMT1. Herein, we report the results of an in-depth and critical comparison of ligand- and structure-based de novo design of screening libraries focused on DNMT1. The newly designed chemical libraries focused on DNMT1 are freely available on GitHub.

Image-based molecular representation learning for drug development: a survey

Artificial intelligence (AI) powered drug development has received remarkable attention in recent years. It addresses the limitations of traditional experimental methods that are costly and time-consuming. While there have been many surveys attempting to summarize related research, they only focus on general AI or specific aspects such as natural language processing and graph neural network. Considering the rapid advance on computer vision, using the molecular image to enable AI appears to be a more intuitive and effective approach since each chemical substance has a unique visual representation. In this paper, we provide the first survey on image-based molecular representation for drug development. The survey proposes a taxonomy based on the learning paradigms in computer vision and reviews a large number of corresponding papers, highlighting the contributions of molecular visual representation in drug development. Besides, we discuss the applications, limitations and future directions in the field. We hope this survey could offer valuable insight into the use of image-based molecular representation learning in the context of drug development.

A terpenoids database with the chemical content as a novel agronomic trait

Natural products play a pivotal role in drug discovery, and the richness of natural products, albeit significantly influenced by various environmental factors, is predominantly determined by intrinsic genetics of a series of enzymatic reactions and produced as secondary metabolites of organisms. Heretofore, few natural product-related databases take the chemical content into consideration as a prominent property. To gain unique insights into the quantitative diversity of natural products, we have developed the first TerPenoids database embedded with Content information (TPCN) with features such as compound browsing, structural search, scaffold analysis, similarity analysis and data download. This database can be accessed through a web-based computational toolkit available at http://www.tpcn.pro/. By conducting meticulous manual searches and analyzing over 10 000 reference papers, the TPCN database has successfully integrated 6383 terpenoids obtained from 1254 distinct plant species. The database encompasses exhaustive details including isolation parts, comprehensive molecule structures, chemical abstracts service registry number (CAS number) and 7508 content descriptions. The TPCN database accentuates both the qualitative and quantitative dimensions as invaluable phenotypic characteristics of natural products that have undergone genetic evolution. By acting as an indispensable criterion, the TPCN database facilitates the discovery of drug alternatives with high content and the selection of high-yield medicinal plant species or phylogenetic alternatives, thereby fostering sustainable, cost-effective and environmentally friendly drug discovery in pharmaceutical farming. Database URL: http://www.tpcn.pro/.

One chiral fingerprint to find them all

Molecular fingerprints are indispensable tools in cheminformatics. However, stereochemistry is generally not considered, which is problematic for large molecules which are almost all chiral. Herein we report MAP4C, a chiral version of our previously reported fingerprint MAP4, which lists MinHashes computed from character strings containing the SMILES of all pairs of circular substructures up to a diameter of four bonds and the shortest topological distance between their central atoms. MAP4C includes the Cahn-Ingold-Prelog (CIP) annotation (R, S, r or s) whenever the chiral atom is the center of a circular substructure, a question mark for undefined stereocenters, and double bond cis-trans information if specified. MAP4C performs slightly better than the achiral MAP4, ECFP and AP fingerprints in non-stereoselective virtual screening benchmarks. Furthermore, MAP4C distinguishes between stereoisomers in chiral molecules from small molecule drugs to large natural products and peptides comprising thousands of diastereomers, with a degree of distinction smaller than between structural isomers and proportional to the number of chirality changes. Due to its excellent performance across diverse molecular classes and its ability to handle stereochemistry, MAP4C is recommended as a generally applicable chiral molecular fingerprint. SCIENTIFIC CONTRIBUTION: The ability of our chiral fingerprint MAP4C to handle stereoisomers from small molecules to large natural products and peptides is unprecedented and opens the way for cheminformatics to include stereochemistry as an important molecular parameter across all fields of molecular design.

Discovery of a SUCNR1 antagonist for potential treatment of diabetic nephropathy: In silico and in vitro studies

Diabetic nephropathy (DN) is one of the most severe complications of diabetes mellitus. Succinate Receptor 1 (SUCNR1), a member of the G-protein-coupled receptor (GPCR) family, represents a potential target for treatment of DN. Here, utilizing multi-strategy in silico virtual screening methods containing AlphaFold2 modelling, molecular dynamics (MD) simulation, ligand-based pharmacophore screening, molecular docking and machine learning-based similarity clustering, we successfully identified a novel antagonist of SUCNR1, AK-968/12117473 (Cpd3). Through extensive in vitro experiments, including dual-luciferase reporter assay, cellular thermal shift assay, immunofluorescence, and western blotting, we substantiated that Cpd3 could specifically target SUCNR1, inhibit the activation of NF-κB pathway, and ameliorate epithelial-mesenchymal transition (EMT) and extracellular matrix (ECM) deposition in renal tubular epithelial cells (NRK-52E) under high glucose conditions. Further in silico simulations revealed the molecular basis of the SUCNR1-Cpd3 interaction, and the in vitro metabolic stability assay indicated favorable drug-like pharmacokinetic properties of Cpd3. This work not only successfully pinpointed Cpd3 as a specific antagonist of SUCNR1 to serve as a promising candidate in the realm of therapeutic interventions for DN, but also provides a paradigm of dry-wet combined discovery strategies for GPCR-based therapeutics.

Soft-sphere continuum solvation models for nonaqueous solvents

Solvation effects profoundly influence the characteristics and behavior of chemical systems in liquid solutions. The interaction between solute and solvent molecules intricately impacts solubility, reactivity, stability, and various chemical processes. Continuum solvation models gained prominence in quantum chemistry by implicitly capturing these interactions and enabling efficient investigations of diverse chemical systems in solution. In comparison, continuum solvation models in condensed matter simulation are very recent. Among these, the self-consistent continuum solvation (SCCS) and the soft-sphere continuum solvation models (SSCS) have been among the first to be successfully parameterized and extended to model periodic systems in aqueous solutions and electrolytes. As most continuum approaches, these models depend on a number of parameters that are linked to experimental or theoretical properties of the solvent, or that can be tuned based on reference data. Here, we present a systematic parameterization of the SSCS model for over 100 nonaqueous solvents. We validate the model's efficacy across diverse solvent environments by leveraging experimental solvation-free energies and partition coefficients from comprehensive databases. The average root means square error over all the solvents was calculated as 0.85 kcal/mol which is below the chemical accuracy (1 kcal/mol). Similarly to what has been reported by Hille et al. (J. Chem. Phys. 2019, 150, 041710.) for the SCCS model, a single-parameter model accurately reproduces experimental solvation energies, showcasing the transferability and predictive power of these continuum approaches. Our findings underscore the potential for a unified approach to predict solvation properties, paving the way for enhanced computational studies across various chemical environments.

Task-Similarity is a Crucial Factor for Few-Shot Meta-Learning of Structure-Activity Relationships

Machine learning models support computer-aided molecular design and compound optimization. However, the initial phases of drug discovery often face a scarcity of training data for these models. Meta-learning has emerged as a potentially promising strategy, harnessing the wealth of structure-activity data available for known targets to facilitate efficient few-shot model training for the specific target of interest. In this study, we assessed the effectiveness of two different meta-learning methods, namely model-agnostic meta-learning (MAML) and adaptive deep kernel fitting (ADKF), specifically in the regression setting. We investigated how factors such as dataset size and the similarity of training tasks impact predictability. The results indicate that ADKF significantly outperformed both MAML and a single-task baseline model on the inhibition data. However, the performance of ADKF varied across different test tasks. Our findings suggest that considerable enhancements in performance can be anticipated primarily when the task of interest is similar to the tasks incorporated in the meta-learning process.

Multiquery Similarity Searching Models: An Alternative Approach for Predicting Hemolytic Activity from Peptide Sequence

The desirable pharmacological properties and a broad number of therapeutic activities have made peptides promising drugs over small organic molecules and antibody drugs. Nevertheless, toxic effects, such as hemolysis, have hampered the development of such promising drugs. Hence, a reliable computational tool to predict peptide hemolytic toxicity is enormously useful before synthesis and experimental evaluation. Currently, four web servers that predict hemolytic activity using machine learning (ML) algorithms are available; however, they exhibit some limitations, such as the need for a reliable negative set and limited application domain. Hence, we developed a robust model based on a novel theoretical approach that combines network science and a multiquery similarity searching (MQSS) method. A total of 1152 initial models were constructed from 144 scaffolds generated in a previous report. These were evaluated on external data sets, and the best models were fused and improved. Our best MQSS model I1 outperformed all state-of-the-art ML-based models and was used to characterize the prevalence of hemolytic toxicity on therapeutic peptides. Based on our model's estimation, the number of hemolytic peptides might be 3.9-fold higher than the reported.

graphLambda: Fusion Graph Neural Networks for Binding Affinity Prediction

Predicting the binding affinity of protein-ligand complexes is crucial for computer-aided drug discovery (CADD) and the identification of potential drug candidates. The deep learning-based scoring functions have emerged as promising predictors of binding constants. Building on recent advancements in graph neural networks, we present graphLambda for protein-ligand binding affinity prediction, which utilizes graph convolutional, attention, and isomorphism blocks to enhance the predictive capabilities. The graphLambda model exhibits superior performance across CASF16 and CSAR HiQ NRC benchmarks and demonstrates robustness with respect to different types of train-validation set partitions. The development of graphLambda underscores the potential of graph neural networks in advancing binding affinity prediction models, contributing to more effective CADD methodologies.

Chemoinformatic approaches for navigating large chemical spaces

INTRODUCTION

Large chemical spaces (CSs) include traditional large compound collections, combinatorial libraries covering billions to trillions of molecules, DNA-encoded chemical libraries comprising complete combinatorial CSs in a single mixture, and virtual CSs explored by generative models. The diverse nature of these types of CSs require different chemoinformatic approaches for navigation.

AREAS COVERED

An overview of different types of large CSs is provided. Molecular representations and similarity metrics suitable for large CS exploration are discussed. A summary of navigation of CSs in generative models is provided. Methods for characterizing and comparing CSs are discussed.

EXPERT OPINION

The size of large CSs might restrict navigation to specialized algorithms and limit it to considering neighborhoods of structurally similar molecules. Efficient navigation of large CSs not only requires methods that scale with size but also requires smart approaches that focus on better but not necessarily larger molecule selections. Deep generative models aim to provide such approaches by implicitly learning features relevant for targeted biological properties. It is unclear whether these models can fulfill this ideal as validation is difficult as long as the covered CSs remain mainly virtual without experimental verification.

Are We Justified in Modeling Human Exposure to Chlorinated Paraffin Mixtures Using the Average Properties of Congeners and Homologues?

Concern over human exposure to chlorinated paraffin (CP) mixtures keeps increasing. The absence of a comprehensive understanding of how human exposure varies with the physicochemical properties of CP constituents has hindered the ability to determine at what level of aggregation exposure to CPs should be assessed. We answer this question by comparing exposure predicted with either a "complex" method that utilizes isomer-specific properties or "simplified" methods that rely on median properties of congener, homologue, or short-/medium-/long-chain CP groups. Our results demonstrate the wide range of physicochemical properties across CP mixtures and their dependence on molecular structures. Assuming unit emissions in the environment, these variances translate into an extensive disparity in whole-body concentrations predicted for different isomers, spanning ∼11 orders of magnitude. CPs with 13-19 carbons and 6-10 chlorines exhibit the highest human exposure potential, primarily owing to moderate to high hydrophobicity and slow environmental degradation and biotransformation. Far-field exposure is dominant for most CP constituents. Our study underscores that using average properties of congener, homologue, or S/M/LCCP groups yields results that are consistent with those derived from isomer-based modeling, thus offering an efficient and practical framework for future risk assessments and human exposure studies of CPs and other complex chemical mixtures.

FlowSOM clustering - A novel pattern recognition approach for water research: Application to a hyper-arid coastal aquifer system

Monitoring and understanding of water resources have become essential in designing effective and sustainable management strategies to overcome the growing water quality challenges. In this context, the utilization of unsupervised learning techniques for evaluating environmental tracers has facilitated the exploration of sources and dynamics of groundwater systems through pattern recognition. However, conventional techniques may overlook spatial and temporal non-linearities present in water research data. This paper introduces the adaptation of FlowSOM, a pioneering approach that combines self-organizing maps (SOM) and minimal spanning trees (MST), with the fast-greedy network clustering algorithm to unravel intricate relationships within multivariate water quality datasets. By capturing connections within the data, this ensemble tool enhances clustering and pattern recognition. Applied to the complex water quality context of the hyper-arid transboundary Caplina/Concordia coastal aquifer system (Peru/Chile), the FlowSOM network and clustering yielded compelling results in pattern recognition of the aquifer salinization. Analyzing 143 groundwater samples across eight variables, including major ions, the approach supports the identification of distinct clusters and connections between them. Three primary sources of salinization were identified: river percolation, slow lateral aquitard recharge, and seawater intrusion. The analysis demonstrated the superiority of FlowSOM clustering over traditional techniques in the case study, producing clusters that align more closely with the actual hydrogeochemical pattern. The outcomes broaden the utilization of multivariate analysis in water research, presenting a comprehensive approach to support the understanding of groundwater systems.

Multi-method computational evaluation of the inhibitors against leucine-rich repeat kinase 2 G2019S mutant for Parkinson's disease

Leucine-rich repeat kinase 2 G2019S mutant (LRRK2 G2019S) is a potential target for Parkinson's disease therapy. In this work, the computational evaluation of the LRRK2 G2019S inhibitors was conducted via a combined approach which contains a preliminary screening of a large database of compounds via similarity and pharmacophore, a secondary selection via structure-based affinity prediction and molecular docking, and a rescoring treatment for the final selection. MD simulations and MM/GBSA calculations were performed to check the agreement between different prediction methods for these inhibitors. 331 experimental ligands were collected, and 170 were used to build the structure-activity relationship. Eight representative ligand structural models were employed in similarity searching and pharmacophore screening over 14 million compounds. The process for selecting proper molecular descriptors provides a successful sample which can be used as a general strategy in QSAR modelling. The rescoring used in this work presents an alternative useful treatment for ranking and selection.

Computational modeling approaches for developing a synergistic effect prediction model of estrogen agonistic activity

The concerns regarding the potential health threats caused by estrogenic endocrine-disrupting chemicals (EDCs) and their mixtures manufactured by the chemical industry are increasing worldwide. Conventional experimental tests for understanding the estrogenic activity of mixtures are expensive and time-consuming. Although non-testing methods using computational modeling approaches have been developed to reduce the number of traditional tests, they are unsuitable for predicting synergistic effects because current prediction models consider only a single chemical. Thus, the development of predictive models is essential for predicting the mixture toxicity, including chemical interactions. However, selecting suitable computational modeling approaches to develop a high-performance prediction model requires considerable time and effort. In this study, we provide a suitable computational approach to develop a predictive model for the synergistic effects of estrogenic activity. We collected datasets on mixture toxicity based on the synergistic effect of estrogen agonistic activity in binary mixtures. Using the model deviation ratio approach, we classified the labels of the binary mixtures as synergistic or non-synergistic effects. We assessed five molecular descriptors, four machine learning-based algorithms, and a deep learning-based algorithm to provide a suitable computational modeling approach. Compared with other modeling approaches, the prediction model using the deep learning-based algorithm and chemical-protein network descriptors exhibited the best performance in predicting the synergistic effects. In conclusion, we developed a new high-performance binary classification model using a deep neural network and chemical-protein network-based descriptors. The developed model will be helpful for the preliminary screening of the synergistic effects of binary mixtures during the development process of chemical products.

Investigation of Zika virus methyl transferase inhibitors using steered molecular dynamics

Zika virus (ZIKV) spread is considered a major public health threat by the World Health Organization (WHO). There are no vaccines or drugs available to control the infection of the Zika virus, therefore a highly effective medicinal molecule is urgently required. In this study, a computationally intensive investigation was performed to identify a potent natural compound that could inhibit the ZIKV NS5 methyltransferase. This research approach is based on target-based drug identification principles where the native inhibitor SAH (S-adenosylhomocysteine) of ZIKV NS5 methyltransferase was selected as a reference. High-throughput virtual screening and tanimoto similarity coefficient were applied to the natural compound library for ranking the potential candidates. The top five compounds were selected for interaction analysis, MD simulation, total binding free energy through MM/GBSA, and steered MD simulation. Among these compounds, Adenosine 5'-monophosphate monohydrate, Tubercidin, and 5-Iodotubercidin showed stable binding to the protein compared to the native compound, SAH. These three compounds also showed less fluctuations in RMSF in contrast to native compound. Additionally, the same interacting residues observed in SAH also made strong interactions with these three compounds. Adenosine 5'-monophosphate monohydrate and 5-Iodotubercidin had greater total binding free energies than the reference ligand. Moreover, the dissociation resistance of all three compounds was equivalent to that of the reference ligand. This study suggested binding properties of three-hit compounds that could be used to develop drugs against Zika virus infections.Communicated by Ramaswamy H. Sarma.

Sequence-Structure Analysis Unlocking the Potential Functional Application of the Local 3D Motifs of Plant-Derived Diterpene Synthases

Plant-derived diterpene synthases (PdiTPSs) play a critical role in the formation of structurally and functionally diverse diterpenoids. However, the specificity or functional-related features of PdiTPSs are not well understood. For a more profound insight, we collected, constructed, and curated 199 functionally characterized PdiTPSs and their corresponding 3D structures. The complex correlations among their sequences, domains, structures, and corresponding products were comprehensively analyzed. Ultimately, our focus narrowed to the geometric arrangement of local structures. We found that local structural alignment can rapidly localize product-specific residues that have been validated by mutagenesis experiments. Based on the 3D motifs derived from the residues around the substrate, we successfully searched diterpene synthases (diTPSs) from the predicted terpene synthases and newly characterized PdiTPSs, suggesting that the identified 3D motifs can serve as distinctive signatures in diTPSs (I and II class). Local structural analysis revealed the PdiTPSs with more conserved amino acid residues show features unique to class I and class II, whereas those with fewer conserved amino acid residues typically exhibit product diversity and specificity. These results provide an attractive method for discovering novel or functionally equivalent enzymes and probing the product specificity in cases where enzyme characterization is limited.

Discovery of Novel Inhibitors of BRD4 for Treating Prostate Cancer: A Comprehensive Case Study for Considering Water Networks in Virtual Screening and Drug Design

Androgen receptor (AR) is the primary target for treating prostate cancer (PCa), which inevitably progresses due to drug-resistant mutations. Bromodomain-containing protein 4 (BRD4) has been a new potential drug target for PCa treatment. Herein, we report the rational design and discovery of novel BRD4 inhibitors through computer-aided drug design (CADD), and a hit compound SQ-1 (IC50 = 676 nM) was identified by structure-based virtual screening (SBVS) with the conserved water network. To optimize the structure of SQ-1, the free energy landscape was constructed, and the binding mechanism was explored by characterizing the water profile and the dissociation mechanism. Finally, the compound SQ-17 with improved inhibitory activity (IC50 < 100 nM) was discovered, which showed potent antiproliferative activity against LNCaP. These data highlighted a successful attempt to identify and optimize a small molecule by comprehensive CADD application and provided essential clues for developing novel therapeutics for PCa treatment.

Automated Virtual Screening

Computational methods in modern drug discovery have become ubiquitous, with methods that cover most of the discovery stages: from hit finding and lead identification to lead optimization. The overall aim of these computational methods is to obtain a more efficient discovery process, by reducing the number of "wet" experiments required to produce therapeutics that have higher probability of succeeding in clinical development and subsequently benefitting end patients by developing highly effective therapeutics having minimal side effects. Virtual Screening is usually applied at the early stage of drug discovery, looking to find chemical matter having desired properties, such as molecular shape, electrostatics, and pharmacophores at desired three-dimensional positions. The aim of this stage is to search in a wide chemical space, including chemistry available from commercial suppliers and virtual databases of predicted reaction products, to identify molecules that would exert a particular biochemical response. This initial stage of the discovery process is very important since the subsequent stages will use the initial chemical motifs that have been found at the hit finding stage, and therefore the most suitable the compound is found, the more likely it is that subsequent stages will be successful and less time and resource consuming. This chapter provides a summary of various Virtual Screening methods, including shape match and molecular docking, and these methods are used in a Virtual Screening workflow that is provided as an example which is described to be run automatically in cloud resources. This automatic in-depth exploration of the chemical space using validated Virtual Screening methods can lead to a more streamlined and efficient discovery process, aiming to deliver chemical matter of high quality and maximizing the required biological effects while minimizing adverse effects. Surely, Virtual Screening pipelines of this nature will continue to play a central role in producing much needed therapeutics for the health challenges of the future.

Transformer-Based Representation of Organic Molecules for Potential Modeling of Physicochemical Properties

In this work, we study the use of three configurations of an autoencoder neural network to process organic substances with the aim of generating meaningful molecular descriptors that can be employed to develop property prediction models. A total of 18,322,500 compounds represented as SMILES strings were used to train the model, demonstrating that a latent space of 24 units is able to adequately reconstruct the data. After AE training, an analysis of the latent space properties in terms of compound similarity was carried out, indicating that this space possesses desired properties for the potential development of models for forecasting physical properties of organic compounds. As a final step, a QSPR model was developed to predict the boiling point of chemical substances based on the AE descriptors. 5276 substances were used for the regression task, and the predictive ability was compared with models available in the literature evaluated on the same database. The final AE model has an overall error of 1.40% (1.39% with augmented SMILES) in the prediction of the boiling temperature, while other models have errors between 2.0 and 3.2%. This shows that the SMILES representation is comparable and even outperforms the state-of-the-art representations widely used in the literature.

Multimodal learning in clinical proteomics: enhancing antimicrobial resistance prediction models with chemical information

MOTIVATION

Large-scale clinical proteomics datasets of infectious pathogens, combined with antimicrobial resistance outcomes, have recently opened the door for machine learning models which aim to improve clinical treatment by predicting resistance early. However, existing prediction frameworks typically train a separate model for each antimicrobial and species in order to predict a pathogen's resistance outcome, resulting in missed opportunities for chemical knowledge transfer and generalizability.

RESULTS

We demonstrate the effectiveness of multimodal learning over proteomic and chemical features by exploring two clinically relevant tasks for our proposed deep learning models: drug recommendation and generalized resistance prediction. By adopting this multi-view representation of the pathogenic samples and leveraging the scale of the available datasets, our models outperformed the previous single-drug and single-species predictive models by statistically significant margins. We extensively validated the multi-drug setting, highlighting the challenges in generalizing beyond the training data distribution, and quantitatively demonstrate how suitable representations of antimicrobial drugs constitute a crucial tool in the development of clinically relevant predictive models.

AVAILABILITY AND IMPLEMENTATION

The code used to produce the results presented in this article is available at https://github.com/BorgwardtLab/MultimodalAMR.

Anatomy of Potency Predictions Focusing on Structural Analogues with Increasing Potency Differences Including Activity Cliffs

Potency predictions are popular in compound design and optimization but are complicated by intrinsic limitations. Moreover, even for nonlinear methods, activity cliffs (ACs, formed by structural analogues with large potency differences) represent challenging test cases for compound potency predictions. We have devised a new test system for potency predictions, including AC compounds, that is based on partitioned matched molecular pairs (MMP) and makes it possible to monitor prediction accuracy at the level of analogue pairs with increasing potency differences. The results of systematic predictions using different machine learning and control methods on MMP-based data sets revealed increasing prediction errors when potency differences between corresponding training and test compounds increased, including large prediction errors for AC compounds. At the global level, these prediction errors were not apparent due to the statistical dominance of analogue pairs with small potency differences. Test compounds from such pairs were accurately predicted and determined the observed global prediction accuracy. Shapley value analysis, an explainable artificial intelligence approach, was applied to identify structural features determining potency predictions using different methods. The analysis revealed that numerical predictions of different regression models were determined by features that were shared by MMP partner compounds or absent in these compounds, with opposing effects. These findings provided another rationale for accurate predictions of similar potency values for structural analogues and failures in predicting the potency of AC compounds.

Metrics for estimating vapour pressure deviation from ideality in binary mixtures

A novel method is introduced for estimating the degree of interactions occurring between two different compounds in a binary mixture resulting in deviations from ideality as predicted by Raoult's law. Metrics of chemical similarity between binary mixture components were used as descriptors and correlated with the Root-Mean Square Error (RMSE) associated with Raoult's law calculations of total vapour pressure prediction, including Abraham descriptors, sigma moments, and several chemical properties. The best correlation was for a quantitative structure-activity relationship (QSAR) equation using differences in Abraham parameters as descriptors (r2 = 0.7585), followed by a QSAR using differences in COSMO-RS sigma moment descriptors (r2 = 0.7461), and third by a QSAR using differences in the chemical properties of log KAW, melting point, and molecular weight as descriptors (r2 = 0.6878). Of these chemical properties, Δlog KAW had the strongest correlation with deviation from Raoult's law (RMSE) and this property alone resulted in an r2 of 0.6630. These correlations are useful for assessing the expected deviation in Raoult's law estimations of vapour pressures, a key property for estimating inhalation exposure.

Clustering Protein Binding Pockets and Identifying Potential Drug Interactions: A Novel Ligand-Based Featurization Method

Protein-ligand interactions are essential to drug discovery and drug development efforts. Desirable on-target or multitarget interactions are the first step in finding an effective therapeutic, while undesirable off-target interactions are the first step in assessing safety. In this work, we introduce a novel ligand-based featurization and mapping of human protein pockets to identify closely related protein targets and to project novel drugs into a hybrid protein-ligand feature space to identify their likely protein interactions. Using structure-based template matches from PDB, protein pockets are featured by the ligands that bind to their best co-complex template matches. The simplicity and interpretability of this approach provide a granular characterization of the human proteome at the protein-pocket level instead of the traditional protein-level characterization by family, function, or pathway. We demonstrate the power of this featurization method by clustering a subset of the human proteome and evaluating the predicted cluster associations of over 7000 compounds.

Stacked ensemble learning on HaCaT cytotoxicity for skin irritation prediction: A case study on dipterocarpol

Skin irritation is an adverse effect associated with various substances, including chemicals, drugs, or natural products. Dipterocarpol, extracted from Dipterocarpus alatus, contains several skin benefits notably anticancer, wound healing, and antibacterial properties. However, the skin irritation of dipterocarpol remains unassessed. Quantitative structure-activity relationship (QSAR) is a recommended tool for toxicity assessment involving less time, money, and animal testing to access unavailable acute toxicity data. Therefore, our study aimed to develop a highly accurate machine learning-based QSAR model for predicting skin irritation. We utilized a stacked ensemble learning model with 1064 chemicals. We also adhered to the recommendations from the OECD for QSAR validation. Subsequently, we used the proposed model to explore the cytotoxicity of dipterocarpol on keratinocytes. Our findings indicate that the model displayed promising statistical quality in terms of accuracy, precision, and recall in both 10-fold cross-validation and test datasets. Moreover, the model predicted that dipterocarpol does not have skin irritation, which was confirmed by the cell-based assay. In conclusion, our proposed model can be applied for the risk assessment of skin irritation in untested compounds that fall within its applicability domain. The web application of this model is available at https://qsarlabs.com/#stackhacat.

Rationalizing general limitations in assessing and comparing methods for compound potency prediction

Compound potency predictions play a major role in computational drug discovery. Predictive methods are typically evaluated and compared in benchmark calculations that are widely applied. Previous studies have revealed intrinsic limitations of potency prediction benchmarks including very similar performance of increasingly complex machine learning methods and simple controls and narrow error margins separating machine learning from randomized predictions. However, origins of these limitations are currently unknown. We have carried out an in-depth analysis of potential reasons leading to artificial outcomes of potency predictions using different methods. Potency predictions on activity classes typically used in benchmark settings were found to be determined by compounds with intermediate potency close to median values of the compound data sets. The potency of these compounds was consistently predicted with high accuracy, without the need for learning, which dominated the results of benchmark calculations, regardless of the activity classes used. Taken together, our findings provide a clear rationale for general limitations of compound potency benchmark predictions and a basis for the design of alternative test systems for methodological comparisons.

A screen of pharmacologically active compounds to identify modulators of the Adgrg6/Gpr126 signalling pathway in zebrafish embryos

Adhesion G protein-coupled receptors (GPCRs) are an underrepresented class of GPCRs in drug discovery. We previously developed an in vivo drug screening pipeline to identify compounds with agonist activity for Adgrg6 (Gpr126), an adhesion GPCR required for myelination of the peripheral nervous system in vertebrates. The screening assay tests for rescue of an ear defect found in adgrg6tb233c-/- hypomorphic homozygous mutant zebrafish, using the expression of versican b (vcanb) mRNA as an easily identifiable phenotype. In the current study, we used the same assay to screen a commercially available library of 1280 diverse bioactive compounds (Sigma LOPAC). Comparison with published hits from two partially overlapping compound collections (Spectrum, Tocris) confirms that the screening assay is robust and reproducible. Using a modified counter screen for myelin basic protein (mbp) gene expression, we have identified 17 LOPAC compounds that can rescue both inner ear and myelination defects in adgrg6tb233c-/- hypomorphic mutants, three of which (ebastine, S-methylisothiourea hemisulfate, and thapsigargin) are new hits. A further 25 LOPAC hit compounds were effective at rescuing the otic vcanb expression but not mbp. Together, these and previously identified hits provide a wealth of starting material for the development of novel and specific pharmacological modulators of Adgrg6 receptor activity.

Impact of structural similarity on the accuracy of retention time prediction

Quantitative Structure-Retention Relationships offer a valuable tool for de-risking chromatographic methods in relation to newly formed or hypothetical compounds, arising from synthetic processes or formulation activities. They can also be used to identify optimal separation conditions, or in support of structural elucidation. In this contribution, we provide a systematic study of the relationship between the accuracy of the retention model, the size of the training set and its structural similarity to the predicted compound. We compare structural similarity expressed either on a fingerprint basis (e.g., Tanimoto index), or by Euclidean distance calculated from of subset of molecular descriptors. The results presented indicate that accurate and predictive models can be built from a small dataset containing as few as 25 compounds, provided that the training set is structurally similar to the test compound. When the training set contains compounds selected by minimizing the Euclidean distance calculated from 3 descriptors most correlated with the retention time, root mean square error of 0.48 min and correlation coefficient of 0.9464 were observed for the test sets of 104 compounds. Moreover, these models meet the Tropsha predictivity criteria. These findings potentially bring the prediction of retention times within the practical reach of pharmaceutical analysts involved in chromatographic method development. We also present an optimisation approach to select algorithm settings in order to minimize the prediction error and ensure model predictivity.

Evaluating Known Zika Virus NS2B-NS3 Protease Inhibitor Scaffolds via In Silico Screening and Biochemical Assays

The NS2B-NS3 protease (NS2B-NS3pro) is regarded as an interesting molecular target for drug design, discovery, and development because of its essential role in the Zika virus (ZIKV) cycle. Although no NS2B-NS3pro inhibitors have reached clinical trials, the employment of drug-like scaffolds can facilitate the screening process for new compounds. In this study, we performed a combination of ligand-based and structure-based in silico methods targeting two known non-peptide small-molecule scaffolds with micromolar inhibitory activity against ZIKV NS2B-NS3pro by a virtual screening (VS) of promising compounds. Based on these two scaffolds, we selected 13 compounds from an initial library of 509 compounds from ZINC15's similarity search. These compounds exhibited structural modifications that are distinct from previously known compounds yet keep pertinent features for binding. Despite promising outcomes from molecular docking and initial enzymatic assays against NS2B-NS3pro, confirmatory assays with a counter-screening enzyme revealed an artifactual inhibition of the assessed compounds. However, we report two compounds, 9 and 11, that exhibited antiviral properties at a concentration of 50 μM in cellular-based assays. Overall, this study provides valuable insights into the ongoing research on anti-ZIKV compounds to facilitate and improve the development of new inhibitors.

De novo drug design based on patient gene expression profiles via deep learning

Computational de novo drug design is a challenging issue in medicine, and it is desirable to consider all of the relevant information of the biological systems in a disease state. Here, we propose a novel computational method to generate drug candidate molecular structures from patient gene expression profiles via deep learning, which we call DRAGONET. Our model can generate new molecules that are likely to counteract disease-specific gene expression patterns in patients, which is made possible by exploring the latent space constructed by a transformer-based variational autoencoder and integrating the substructures of disease-correlated molecules. We applied DRAGONET to generate drug candidate molecules for gastric cancer, atopic dermatitis, and Alzheimer's disease, and demonstrated that the newly generated molecules were chemically similar to registered drugs for each disease. This approach is applicable to diseases with unknown therapeutic target proteins and will make a significant contribution to the field of precision medicine.

Artificial Intelligence in Decrypting Cytoprotective Activity under Oxidative Stress from Molecular Structure

Artificial intelligence (AI) is widely explored nowadays, and it gives opportunities to enhance classical approaches in QSAR studies. The aim of this study was to investigate the cytoprotective activity parameter under oxidative stress conditions for indole-based structures, with the ultimate goal of developing AI models capable of predicting cytoprotective activity and generating novel indole-based compounds. We propose a new AI system capable of suggesting new chemical structures based on some known cytoprotective activity. Cytoprotective activity prediction models, employing algorithms such as random forest, decision tree, support vector machines, K-nearest neighbors, and multiple linear regression, were built, and the best (based on quality measurements) was used to make predictions. Finally, the experimental evaluation of the computational results was undertaken in vitro. The proposed methodology resulted in the creation of a library of new indole-based compounds with assigned cytoprotective activity. The other outcome of this study was the development of a validated predictive model capable of estimating cytoprotective activity to a certain extent using molecular structure as input, supported by experimental confirmation.

Identification of Macrocyclic Peptide Families from Combinatorial Libraries Containing Noncanonical Amino Acids Using Cheminformatics and Bioinformatics Inspired Clustering

In the past decade, macrocyclic peptides gained increasing interest as a new therapeutic modality to tackle intracellular and extracellular therapeutic targets that had been previously classified as "undruggable". Several technological advances have made discovering macrocyclic peptides against these targets possible: 1) the inclusion of noncanonical amino acids (NCAAs) into mRNA display, 2) increased availability of next generation sequencing (NGS), and 3) improvements in rapid peptide synthesis platforms. This type of directed-evolution based screening can produce large numbers of potential hit sequences given that DNA sequencing is the functional output of this platform. The current standard for selecting hit peptides from these selections for downstream follow-up relies on the frequency counting and sorting of unique peptide sequences which can result in the generation of false negatives due to technical reasons including low translation efficiency or other experimental factors. To overcome our inability to detect weakly enriched peptide sequences among our large data sets, we wanted to develop a clustering method that would enable the identification of peptide families. Unfortunately, utilizing traditional clustering algorithms, such as ClustalW, is not possible for this technology due to the incorporation of NCAAs in these libraries. Therefore, we developed a new atomistic clustering method with a Pairwise Aligned Peptide (PAP) chemical similarity metric to perform sequence alignments and identify macrocyclic peptide families. With this method, low enriched peptides, including isolated sequences (singletons), can now be clustered into families providing a comprehensive analysis of NGS data resulting from macrocycle discovery selections. Additionally, upon identification of a hit peptide with the desired activity, this clustering algorithm can be used to identify derivatives from the initial data set for structure-activity relationship (SAR) analysis without requiring additional selection experiments.

Metabolic profiling of Chimonanthus grammatus via UHPLC-HRMS-MS with computer-assisted structure elucidation and its antimicrobial activity

Chimonanthus grammatus is used as Hakka traditional herb to treat cold, flu, etc. So far, the phytochemistry and antimicrobial compounds have not been well investigated. In this study, the orbitrap-ion trap MS was used to characterize its metabolites, combined with a computer-assisted structure elucidation method, and the antimicrobial activities were assessed by a broth dilution method against 21 human pathogens, as well as the bioassay-guided purification work to clarify its main antimicrobial compounds. A total of 83 compounds were identified with their fragmentation patterns, including terpenoids, coumarins, flavonoids, organic acids, alkaloids, and others. The plant extracts can strongly inhibit the growth of three Gram-positive and four Gram-negative bacteria, and nine active compounds were bioassay-guided isolated, including homalomenol C, jasmonic acid, isofraxidin, quercitrin, stigmasta-7,22-diene-3β,5α,6α-triol, quercetin, 4-hydroxy-1,10-secocadin-5-ene-1,10-dione, kaempferol, and E-4-(4,8-dimethylnona-3,7-dienyl)furan-2(5H)-one. Among them, isofraxidin, kaempferol, and quercitrin showed significant activity against planktonic Staphylococcus aureus (IC₅₀ = 13.51, 18.08 and 15.86 µg/ml). Moreover, their antibiofilm activities of S. aureus (BIC₅₀ = 15.43, 17.31, 18.86 µg/ml; BEC₅₀ = 45.86, ≥62.50, and 57.62 µg/ml) are higher than ciprofloxacin. The results demonstrated that the isolated antimicrobial compounds played the key role of this herb in combating microbes and provided benefits for its development and quality control, and the computer-assisted structure elucidation method was a powerful tool for chemical analysis, especially for distinguishing isomers with similar structures, which can be used for other complex samples.

Prediction of KRASG12C inhibitors using conjoint fingerprint and machine learning-based QSAR models

Kirsten rat sarcoma virus G12C (KRAS^G12C) is the major protein mutation associated with non-small cell lung cancer (NSCLC) severity. Inhibiting KRAS^G12C is therefore one of the key therapeutic strategies for NSCLC patients. In this paper, a cost-effective data driven drug design employing machine learning-based quantitative structure-activity relationship (QSAR) analysis was built for predicting ligand affinities against KRAS^G12C protein. A curated and non-redundant dataset of 1033 compounds with KRAS^G12C inhibitory activity (pIC₅₀) was used to build and test the models. The PubChem fingerprint, Substructure fingerprint, Substructure fingerprint count, and the conjoint fingerprint-a combination of PubChem fingerprint and Substructure fingerprint count-were used to train the models. Using comprehensive validation methods and various machine learning algorithms, the results clearly showed that the XGBoost regression (XGBoost) achieved the highest performance in term of goodness of fit, predictivity, generalizability and model robustness (R² = 0.81, Q²_CV = 0.60, Q²_Ext = 0.62, R² - Q²_Ext = 0.19, R²_Y-Random = 0.31 ± 0.03, Q²_Y-Random = -0.09 ± 0.04). The top 13 molecular fingerprints that correlated with the predicted pIC₅₀ values were SubFPC274 (aromatic atoms), SubFPC307 (number of chiral-centers), PubChemFP37 (≥1 Chlorine), SubFPC18 (Number of alkylarylethers), SubFPC1 (number of primary carbons), SubFPC300 (number of 1,3-tautomerizables), PubChemFP621 (N-C:C:C:N structure), PubChemFP23 (≥1 Fluorine), SubFPC2 (number of secondary carbons), SubFPC295 (number of C-ONS bonds), PubChemFP199 (≥4 6-membered rings), PubChemFP180 (≥1 nitrogen-containing 6-membered ring), and SubFPC180 (number of tertiary amine). These molecular fingerprints were virtualized and validated using molecular docking experiments. In conclusion, this conjoint fingerprint and XGBoost-QSAR model demonstrated to be useful as a high-throughput screening tool for KRAS^G12C inhibitor identification and drug design.

Systematic investigation of the multi-scale mechanisms of herbal medicine on treating ventricular remodeling: Theoretical and experimental studies

BACKGROUND

To explore the underlying molecule mechanism of herbal medicine in preventing ventricular remodeling (VR), we take a herbal formula that is clinically effective for preventing VR as an example, which composed of Pachyma hoelen Rumph, Atractylodes macrocephala Koidz., Cassia Twig and Licorice. Due to multi-components and multi-targets in herbal medicine, it is extremely difficult to systematically explain its mechanisms of action.

METHODS

An innovative systematic investigation framework which combines with pharmacokinetic screening, target fishing, network pharmacology, DeepDDI algorithm, computational chemistry, molecular thermodynamics, in vivo and in vitro experiments was performed for deciphering the underlying molecular mechanisms of herbal medicine for treating VR.

RESULTS

ADME screening and SysDT algorithm determined 75 potentially active compounds and 109 corresponding targets. Then, systematic analysis of networks reveals the crucial active ingredients and key targets in herbal medicine. Additionally, transcriptomic analysis identifies 33 key regulators during VR progression. Moreover, PPI network and biological function enrichment present four crucial signaling pathways, i.e. NF-κB and TNF, PI3K-AKT and C-type lectin receptor signaling pathways involved in VR. Besides, both molecular experiments at animal and cell levels reveal the beneficial effect of herbal medicine on preventing VR. Finally, MD simulations and binding free energy validate the reliability of drug-target interactions.

CONCLUSION

Our novelty is to build a systematic strategy which combines various theoretical methods combined with experimental approaches. This strategy provides a deep understanding for the study of molecular mechanisms of herbal medicine on treating diseases from systematic level, and offers a new idea for modern medicine to explore drug interventions for complex diseases as well.

Comprehensive assessment of nine target prediction web services: which should we choose for target fishing?

Identification of potential targets for known bioactive compounds and novel synthetic analogs is of considerable significance. In silico target fishing (TF) has become an alternative strategy because of the expensive and laborious wet-lab experiments, explosive growth of bioactivity data and rapid development of high-throughput technologies. However, these TF methods are based on different algorithms, molecular representations and training datasets, which may lead to different results when predicting the same query molecules. This can be confusing for practitioners in practical applications. Therefore, this study systematically evaluated nine popular ligand-based TF methods based on target and ligand-target pair statistical strategies, which will help practitioners make choices among multiple TF methods. The evaluation results showed that SwissTargetPrediction was the best method to produce the most reliable predictions while enriching more targets. High-recall similarity ensemble approach (SEA) was able to find real targets for more compounds compared with other TF methods. Therefore, SwissTargetPrediction and SEA can be considered as primary selection methods in future studies. In addition, the results showed that k = 5 was the optimal number of experimental candidate targets. Finally, a novel ensemble TF method based on consensus voting is proposed to improve the prediction performance. The precision of the ensemble TF method outperforms the individual TF method, indicating that the ensemble TF method can more effectively identify real targets within a given top-k threshold. The results of this study can be used as a reference to guide practitioners in selecting the most effective methods in computational drug discovery.

Seawater intrusion pattern recognition supported by unsupervised learning: A systematic review and application

Seawater intrusion is among the world's leading causes of groundwater contamination, as salty water can affect potable water access, food production, and ecosystem functions. To explore such contamination sources, multivariate analysis supported by unsupervised learning tools has been used for decades to aid in water resource pattern recognition, clustering, and water quality data variability characterization. This study proposes a systematic review of these techniques applied for supporting seawater intrusion identification based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and subsequent bibliometric analysis of 102 coastal hydrogeological studies. The most relevant identified methods, including principal components analysis (PCA), hierarchical clustering analysis, K-means clustering, and self-organizing maps, are explained and applied to a case study. Although 74 % of the studies that applied dimensional reduction methods, such as PCA, associated most of the database variance with the salinization process, 77 % of the studies that applied clustering methods associated at least one water sample cluster with the influence of seawater intrusion. Based on the review and a practical demonstration using the open-source R software platform, recommendations are made regarding data preprocessing, research opportunities, and publishing information necessary to replicate and validate the studies.

Exploration of LPS2 agonist binding modes using the combination of a new hydrophobic scaffold and homology modeling

Lysophosphatidylserine (LysoPS) is an endogenous pan-agonist of three G-protein coupled receptors (GPCRs): LPS₁/GPR34, LPS₂/P2Y₁₀, and LPS₃/GPR174, and we previously reported a series of LysoPS-based agonists of these receptors. Interestingly, we found that LPS₁ agonist activity was very sensitive to structural change at the hydrophobic fatty acid moiety, whereas LPS₂ agonist activity was not. Here, to probe the molecular basis of LPS₂ agonist binding, we developed a new class of hydrophobic fatty acid surrogates having a biphenyl-ether scaffold. The LPS₂ agonist activity of these compounds proved sensitive to molecular modification of the hydrophobic skeleton. Thus, we next constructed an LPS₂ model by homology modeling and docking/molecular dynamics (MD) simulation, and validated it by means of SAR studies together with point mutations of selected receptor amino-acid residues. The putative ligand-binding site of LPS₂ is Γ-shaped, with a hydrophilic site horizontally embedded in the receptor transmembrane helix bundles and a perpendicular hydrophobic groove adjoining transmembrane domains 4 and 5 that is open to the membrane bilayer. The binding poses of LPS₂ agonists to this site are consistent with easy incorporation of various kinds of fatty acid surrogates. Structural development based on this model afforded a series of potent and selective LPS₂ full agonists, which showed enhanced in vitro actin stress fiber formation effect.

Exploring the Potential of Phytocompounds for Targeting Epigenetic Mechanisms in Rheumatoid Arthritis: An In Silico Study Using Similarity Indexing

Finding structurally similar compounds in compound databases is highly efficient and is widely used in present-day drug discovery methodology. The most-trusted and -followed similarity indexing method is Tanimoto similarity indexing. Epigenetic proteins like histone deacetylases (HDACs) inhibitors are traditionally used to target cancer, but have only been investigated very recently for their possible effectiveness against rheumatoid arthritis (RA). The synthetic drugs that have been identified and used for the inhibition of HDACs include SAHA, which is being used to inhibit the activity of HDACs of different classes. SAHA was chosen as a compound of high importance as it is reported to inhibit the activity of many HDAC types. Similarity searching using the UNPD database as a reference identified aglaithioduline from the Aglaia leptantha compound as having a ~70% similarity of molecular fingerprints with SAHA, based on the Tanimoto indexing method using ChemmineR. Aglaithioduline is abundantly present in the shell and fruits of A. leptantha. In silico studies with aglaithioduline were carried out against the HDAC8 protein target and showed a binding affinity of -8.5 kcal mol. The complex was further subjected to molecular dynamics simulation using Gromacs. The RMSD, RMSF, compactness and SASA plots of the target with aglaithioduline, in comparison with the co-crystallized ligand (SAHA) system, showed a very stable configuration. The results of the study are supportive of the usage of A. leptantha and A. edulis in Indian traditional medicine for the treatment of pain-related ailments similar to RA. Our study therefore calls for further investigation of A. leptantha and A. edulis for their potential use against RA by targeting epigenetic changes, using in vivo and in vitro studies.

Hyper-Mol: Molecular Representation Learning via Fingerprint-Based Hypergraph

With the development of artificial intelligence (AI) in the field of drug design and discovery, learning informative representations of molecules is becoming crucial for those AI-driven tasks. In recent years, the graph neural networks (GNNs) have emerged as a preferred choice of deep learning architecture and have been successfully applied to molecular representation learning (MRL). Up-to-date MRL methods directly apply the message passing mechanism on the atom-level attributes (i.e., atoms and bonds) of molecules. However, they neglect latent yet significant hyperstructured knowledge, such as the information of pharmacophore or functional class. Hence, in this paper, we propose Hyper-Mol, a new MRL framework that applies GNNs to encode hypergraph structures of molecules via fingerprint-based features. Hyper-Mol explores the hyperstructured knowledge and the latent relationships of the fingerprint substructures from a hypergraph perspective. The molecular hypergraph generation algorithm is designed to depict the hyperstructured information with the physical and chemical characteristics of molecules. Thus, the fingerprint-level message passing process can encode both the intra-structured and inter-structured information of fingerprint substructures according to the molecular hypergraphs. We evaluate Hyper-Mol on molecular property prediction tasks, and the experimental results on real-world benchmarks show that Hyper-Mol can learn comprehensive hyperstructured knowledge of molecules and is superior to the state-of-the-art baselines.

Computer-Aided Drug Design towards New Psychotropic and Neurological Drugs

Central nervous system (CNS) disorders are a therapeutic area in drug discovery where demand for new treatments greatly exceeds approved treatment options. This is complicated by the high failure rate in late-stage clinical trials, resulting in exorbitant costs associated with bringing new CNS drugs to market. Computer-aided drug design (CADD) techniques minimise the time and cost burdens associated with drug research and development by ensuring an advantageous starting point for pre-clinical and clinical assessments. The key elements of CADD are divided into ligand-based and structure-based methods. Ligand-based methods encompass techniques including pharmacophore modelling and quantitative structure activity relationships (QSARs), which use the relationship between biological activity and chemical structure to ascertain suitable lead molecules. In contrast, structure-based methods use information about the binding site architecture from an established protein structure to select suitable molecules for further investigation. In recent years, deep learning techniques have been applied in drug design and present an exciting addition to CADD workflows. Despite the difficulties associated with CNS drug discovery, advances towards new pharmaceutical treatments continue to be made, and CADD has supported these findings. This review explores various CADD techniques and discusses applications in CNS drug discovery from 2018 to November 2022.

Neural Networks in the Design of Molecules with Affinity to Selected Protein Domains

Drug design with machine learning support can speed up new drug discoveries. While current databases of known compounds are smaller in magnitude (approximately 108), the number of small drug-like molecules is estimated to be between 1023 and 1060. The use of molecular docking algorithms can help in new drug development by sieving out the worst drug-receptor complexes. New chemical spaces can be efficiently searched with the application of artificial intelligence. From that, new structures can be proposed. The research proposed aims to create new chemical structures supported by a deep neural network that will possess an affinity to the selected protein domains. Transferring chemical structures into SELFIES codes helped us pass chemical information to a neural network. On the basis of vectorized SELFIES, new chemical structures can be created. With the use of the created neural network, novel compounds that are chemically sensible can be generated. Newly created chemical structures are sieved by the quantitative estimation of the drug-likeness descriptor, Lipinski's rule of 5, and the synthetic Bayesian accessibility classifier score. The affinity to selected protein domains was verified with the use of the AutoDock tool. As per the results, we obtained the structures that possess an affinity to the selected protein domains, namely PDB IDs 7NPC, 7NP5, and 7KXD.

Substantiate a read-across hypothesis by using transcriptome data-A case study on volatile diketones

This case study explores the applicability of transcriptome data to characterize a common mechanism of action within groups of short-chain aliphatic α-, β-, and γ-diketones. Human reference in vivo data indicate that the α-diketone diacetyl induces bronchiolitis obliterans in workers involved in the preparation of microwave popcorn. The other three α-diketones induced inflammatory responses in preclinical in vivo animal studies, whereas beta and gamma diketones in addition caused neuronal effects. We investigated early transcriptional responses in primary human bronchiolar (PBEC) cell cultures after 24 h and 72 h of air-liquid exposure. Differentially expressed genes (DEGs) were assessed based on transcriptome data generated with the EUToxRisk gene panel of Temp-O-Seq^®. For each individual substance, genes were identified displaying a consistent differential expression across dose and exposure duration. The log fold change values of the DEG profiles indicate that α- and β-diketones are more active compared to γ-diketones. α-diketones in particular showed a highly concordant expression pattern, which may serve as a first indication of the shared mode of action. In order to gain a better mechanistic understanding, the resultant DEGs were submitted to a pathway analysis using ConsensusPathDB. The four α-diketones showed very similar results with regard to the number of activated and shared pathways. Overall, the number of signaling pathways decreased from α-to β-to γ-diketones. Additionally, we reconstructed networks of genes that interact with one another and are associated with different adverse outcomes such as fibrosis, inflammation or apoptosis using the TRANSPATH-database. Transcription factor enrichment and upstream analyses with the geneXplain platform revealed highly interacting gene products (called master regulators, MRs) per case study compound. The mapping of the resultant MRs on the reconstructed networks, visualized similar gene regulation with regard to fibrosis, inflammation and apoptosis. This analysis showed that transcriptome data can strengthen the similarity assessment of compounds, which is of particular importance, e.g., in read-across approaches. It is one important step towards grouping of compounds based on biological profiles.