Scaffold hopping

AutoDesigner - Core Design, a De Novo Design Algorithm for Chemical Scaffolds: Application to the Design and Synthesis of Novel Selective Wee1 Inhibitors

The hit identification stage of a drug discovery program generally involves the design of novel chemical scaffolds with desired biological activity against the target(s) of interest. One common approach is scaffold hopping, which is the manual design of novel scaffolds based on known chemical matter. One major limitation of this approach is narrow chemical space exploration, which can lead to difficulties in maintaining or improving biological activity, selectivity, and favorable property space. Another limitation is the lack of preliminary structure-activity relationship (SAR) data around these designs, which could lead to selecting suboptimal scaffolds to advance lead optimization. To address these limitations, we propose AutoDesigner - Core Design (CoreDesign), a de novo scaffold design algorithm. Our approach is a cloud-integrated, de novo design algorithm for systematically exploring and refining chemical scaffolds against biological targets of interest. The algorithm designs, evaluates, and optimizes a vast range, from millions to billions, of molecules in silico, following defined project parameters encompassing structural novelty, physicochemical attributes, potency, and selectivity using active-learning FEP. To validate CoreDesign in a real-world drug discovery setting, we applied it to the design of novel, potent Wee1 inhibitors with improved selectivity over PLK1. Starting from a single known ligand and receptor structure, CoreDesign rapidly explored over 23 billion molecules to identify 1,342 novel chemical series with a mean of 4 compounds per scaffold. To rapidly analyze this large amount of data and prioritize chemical scaffolds for synthesis, we utilize t-Distributed Stochastic Neighbor Embedding (t-SNE) plots of in silico properties. The chemical space projections allowed us to rapidly identify a structurally novel 5-5 fused core meeting all the hit-identification requirements. Several compounds were synthesized and assayed from the scaffold, displaying good potency against Wee1 and excellent PLK1 selectivity. Our results suggest that CoreDesign can significantly speed up the hit-identification process and increase the probability of success of drug discovery campaigns by allowing teams to bring forward high-quality chemical scaffolds derisked by the availability of preliminary SAR.

Unlocking comprehensive molecular design across all scenarios with large language model and unordered chemical language

Molecular generation stands at the forefront of AI-driven technologies, playing a crucial role in accelerating the development of small molecule drugs. The intricate nature of practical drug discovery necessitates the development of a versatile molecular generation framework that can tackle diverse drug design challenges. However, existing methodologies often struggle to encompass all aspects of small molecule drug design, particularly those rooted in language models, especially in tasks like linker design, due to the autoregressive nature of large language model-based approaches. To empower a language model for a wider range of molecular design tasks, we introduce an unordered simplified molecular-input line-entry system based on fragments (FU-SMILES). Building upon this foundation, we propose FragGPT, a universal fragment-based molecular generation model. Initially pretrained on extensive molecular datasets, FragGPT utilizes FU-SMILES to facilitate efficient generation across various practical applications, such as de novo molecule design, linker design, R-group exploration, scaffold hopping, and side chain optimization. Furthermore, we integrate conditional generation and reinforcement learning (RL) methodologies to ensure that the generated molecules possess multiple desired biological and physicochemical properties. Experimental results across diverse scenarios validate FragGPT's superiority in generating molecules with enhanced properties and novel structures, outperforming existing state-of-the-art models. Moreover, its robust drug design capability is further corroborated through real-world drug design cases.

Developing Allosteric Inhibitors of SARS-CoV-2 RNA-Dependent RNA Polymerase

The use of Fpocket and virtual screening techniques enabled us to identify potential allosteric druggable pockets within the SARS-CoV-2 RNA-dependent RNA polymerase (RdRp). Of the compounds screened, compound 1 was identified as a promising inhibitor, lowering a SARS-CoV-2 RdRp activity to 57 % in an enzymatic assay at 10 μM concentration. The structure of compound 1 was subsequently optimized in order to preserve or enhance inhibitory activity. This involved the substitution of problematic ester and aromatic nitro groups with more inert functionalities. The N,N'-diphenylurea scaffold with two NH groups was identified as essential for the compound's activity but also exhibited high toxicity in Calu-3 cells. To address this issue, a scaffold hopping approach was employed to replace the urea core with potentially less toxic urea isosteres. This approach yielded several structural analogues with notable activity, specifically 2,2'-bisimidazol (in compound 55 with residual activity RA=42 %) and (1H-imidazol-2-yl)urea (in compounds 59 and 60, with RA=50 and 28 %, respectively). Despite these advances, toxicity remained a major concern. These compounds represent a promising starting point for further structure-activity relationship studies of allosteric inhibitors of SARS-CoV-2 RdRp, with the goal of reducing their cytotoxicity and improving aqueous solubility.

Rutaecarpine-inspired scaffold-hopping strategy and Ullmann cross-coupling based synthetic approach: Identification of pyridopyrimidinone-indole based novel anticancer chemotypes

Natural products as starting templates have shown historically major contribution to development of drugs. Inspired by the structure-function of an anticancer natural alkaloid Rutaecarpine, the Scaffold-hopped Acyclic Analogues of Rutaecarpine (SAAR) with 'N'-atom switch (1°-hop) and ring-opening (2°-hop) were investigated. A new synthetic route was developed for an effective access to the analogues, i.e. 2-indolyl-pyrido[1,2-a]pyrimidinones, which involved preparation of N-Boc-N'-phthaloyltryptamine/mexamine-bromides and pyridopyrmidinon-2-yl triflate, a nickel/palladium-catalysed Ullmann cross-coupling of these bromides and triflate, deprotection of phthalimide followed by N-aroylation, and Boc-deprotection. Fourteen novel SAAR-compounds were prepared, and they showed characteristic antiproliferative activity against various cancer cells. Three most active compounds (11a, 11b, and 11c) exhibited good antiproliferative activity, IC50 7.7-15.8 µM against human breast adenocarcinoma cells (MCF-7), lung cancer cells (A549), and colon cancer cells (HCT-116). The antiproliferative property was also observed in the colony formation assay. The SAAR compound 11b was found to have superior potency than original natural product Rutaecarpine and an anticancer drug 5-FU in antiproliferative activities with relatively lower cytotoxicity towards normal breast epithelial cells (MCF10A) and significantly higher inhibitory effect on cancer cells' migration. The compound 11b was found to possess favourable in silico physicochemical characteristics (lipophilicity-MLOGP, TPSA, and water solubility-ESOL, and others), bioavailability score, and pharmacokinetic properties (GI absorption, BBB non-permeant, P-gp, and CYP2D6). Interestingly, the compound 11b did not show any medicinal chemistry structural alert of PAINS and Brenk filter. The study represents for the first time the successful discovery of new potent anticancer chemotypes using Rutaecarpine natural alkaloid as starting template and reaffirms the significance of natural product-inspired scaffold-hopping technique in drug discovery research.

Scaffold-Hopping Strategies in Aurone Optimization: A Comprehensive Review of Synthetic Procedures and Biological Activities of Nitrogen and Sulfur Analogues

Aurones, particular polyphenolic compounds belonging to the class of minor flavonoids and overlooked for a long time, have gained significative attention in medicinal chemistry in recent years. Indeed, considering their unique and outstanding biological properties, they stand out as an intriguing reservoir of new potential lead compounds in the drug discovery context. Nevertheless, several physicochemical, pharmacokinetic, and pharmacodynamic (P3) issues hinder their progression in more advanced phases of the drug discovery pipeline, making lead optimization campaigns necessary. In this context, scaffold hopping has proven to be a valuable approach in the optimization of natural products. This review provides a comprehensive and updated picture of the scaffold-hopping approaches directed at the optimization of natural and synthetic aurones. In the literature analysis, a particular focus is given to nitrogen and sulfur analogues. For each class presented, general synthetic procedures are summarized, highlighting the key advantages and potential issues. Furthermore, the biological activities of the most representative scaffold-hopped compounds are presented, emphasizing the improvements achieved and the potential for further optimization compared to the aurone class.

Insights into the computer-aided drug design and discovery based on anthraquinone scaffold for cancer treatment: A systematic review

BACKGROUND

In the search for better anticancer drugs, computer-aided drug design (CADD) techniques play an indispensable role in facilitating the lengthy and costly drug discovery process especially when natural products are involved. Anthraquinone is one of the most widely-recognized natural products with anticancer properties. This review aimed to systematically assess and synthesize evidence on the utilization of CADD techniques centered on the anthraquinone scaffold for cancer treatment.

METHODS

The conduct and reporting of this review were done in accordance to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) 2020 guideline. The protocol was registered in the "International prospective register of systematic reviews" database (PROSPERO: CRD42023432904) and also published recently. The search strategy was designed based on the combination of concept 1 "CADD or virtual screening", concept 2 "anthraquinone" and concept 3 "cancer". The search was executed in PubMed, Scopus, Web of Science and MedRxiv on 30 June 2023.

RESULTS

Databases searching retrieved a total of 317 records. After deduplication and applying the eligibility criteria, the final review ended up with 32 articles in which 3 articles were found by citation searching. The CADD methods used in the studies were either structure-based alone (69%) or combined with ligand-based methods via parallel (9%) or sequential (22%) approaches. Molecular docking was performed in all studies, with Glide and AutoDock being the most popular commercial and public software used respectively. Protein data bank was used in most studies to retrieve the crystal structure of the targets of interest while the main ligand databases were PubChem and Zinc. The utilization of in-silico techniques has enabled a deeper dive into the structural, biological and pharmacological properties of anthraquinone derivatives, revealing their remarkable anticancer properties in an all-rounded fashion.

CONCLUSION

By harnessing the power of computational tools and leveraging the natural diversity of anthraquinone compounds, researchers can expedite the development of better drugs to address the unmet medical needs in cancer treatment by improving the treatment outcome for cancer patients.

Can Pretrained Models Really Learn Better Molecular Representations for AI-Aided Drug Discovery?

Self-supervised pretrained models are gaining increasingly more popularity in AI-aided drug discovery, leading to more and more pretrained models with the promise that they can extract better feature representations for molecules. Yet, the quality of learned representations has not been fully explored. In this work, inspired by the two phenomena of Activity Cliffs (ACs) and Scaffold Hopping (SH) in traditional Quantitative Structure-Activity Relationship analysis, we propose a method named Representation-Property Relationship Analysis (RePRA) to evaluate the quality of the representations extracted by the pretrained model and visualize the relationship between the representations and properties. The concepts of ACs and SH are generalized from the structure-activity context to the representation-property context, and the underlying principles of RePRA are analyzed theoretically. Two scores are designed to measure the generalized ACs and SH detected by RePRA, and therefore, the quality of representations can be evaluated. In experiments, representations of molecules from 10 target tasks generated by 7 pretrained models are analyzed. The results indicate that the state-of-the-art pretrained models can overcome some shortcomings of canonical Extended-Connectivity FingerPrints, while the correlation between the basis of the representation space and specific molecular substructures are not explicit. Thus, some representations could be even worse than the canonical fingerprints. Our method enables researchers to evaluate the quality of molecular representations generated by their proposed self-supervised pretrained models. And our findings can guide the community to develop better pretraining techniques to regularize the occurrence of ACs and SH.

Sulfone-based human liver pyruvate kinase inhibitors - Design, synthesis and in vitro bioactivity

Non-alcoholic fatty liver disease (NAFLD) is a prevalent pathological condition characterised by the accumulation of fat in the liver. Almost one-third of the global population is affected by NAFLD, making it a significant health concern. However, despite its prevalence, there is currently no approved drug specifically designed for the treatment of NAFLD. To address this critical gap, researchers have been investigating potential targets for NAFLD drug development. One promising candidate is the liver isoform of pyruvate kinase (PKL). In recent studies, Urolithin C, an allosteric inhibitor of PKL, has emerged as a potential lead compound for therapeutic intervention. Building upon this knowledge, our team has conducted a comprehensive structure-activity relationship of Urolithin C. In this work, we have employed a scaffold-hopping approach, modifying the urolithin structure by replacing the urolithin carbonyl with a sulfone moiety. Our structure-activity relationship analysis has identified the sulfone group as particularly favourable for potent PKL inhibition. Additionally, we have found that the presence of catechol moieties on the two aromatic rings further improves the inhibitory activity. The most promising inhibitor from this new series displayed nanomolar inhibition, boasting an IC50 value of 0.07 μM. This level of potency rivals that of urolithin D and significantly surpasses the effectiveness of urolithin C by an order of magnitude. To better understand the molecular interactions underlying this inhibition, we obtained the crystal structure of one of the inhibitors complexed with PKL. This structural insight served as a valuable reference point, aiding us in the design of inhibitors.

Ligand-based pharmacophore modeling and machine learning for the discovery of potent aurora A kinase inhibitory leads of novel chemotypes

Aurora-A (AURKA) is serine/threonine protein kinase involved in the regulation of numerous processes of cell division. Numerous studies have demonstrated strong association between AURKA and cancer. AURKA is overexpressed in many cancers, such as colon, breast and prostate cancers. Consequently, AURKA has emerged as promising target for therapeutic intervention in cancer management. Herein, we describe a computational workflow for the discovery of novel anti-AURKA inhibitory leads starting with ligand-based assessment of the pharmacophoric space of six diverse sets of inhibitors. Subsequently, machine learning/QSAR modeling was coupled with genetic function algorithm to search for the best possible combination of machine learner, ligand-based pharmacophore(s) and molecular descriptors capable of explaining variation in anti-AURKA bioactivities within a collected list of inhibitors. Two learners succeeded in achieving acceptable structure/activity correlations, namely, random forests and extreme gradient boosting (XGBoost). Three pharmacophores emerged in the successful ML models. These were then used as 3D search queries to mine the National Cancer Institute database for novel anti-AURKA leads. Top-ranking 38 hits were assessed in vitro for their anti-AURKA bioactivities. Among them, three compounds exhibited promising dose-response curves, demonstrating experimental IC50 values ranging from sub-micromolar to low micromolar values. Remarkably, two of these compounds are of novel chemotypes.

Mechanistic Investigation, Wavelength-Dependent Reactivity, and Expanded Reactivity of N-Aryl Azacycle Photomediated Ring Contractions

Under mild blue-light irradiation, α-acylated saturated heterocycles undergo a photomediated one-atom ring contraction that extrudes a heteroatom from the cyclic core. However, for nitrogenous heterocycles, this powerful skeletal edit has been limited to substrates bearing electron-withdrawing substituents on nitrogen. Moreover, the mechanism and wavelength-dependent efficiency of this transformation have remained unclear. In this work, we increased the electron richness of nitrogen in saturated azacycles to improve light absorption and strengthen critical intramolecular hydrogen bonding while enabling the direct installation of the photoreactive handle. As a result, a broadly expanded substrate scope, including underexplored electron-rich substrates and previously unsuccessful heterocycles, has now been achieved. The significantly improved yields and diastereoselectivities have facilitated reaction rate, kinetic isotope effect (KIE), and quenching studies, in addition to the determination of quantum yields. Guided by these studies, we propose a revised ET/PT mechanism for the ring contraction, which is additionally corroborated by computational characterization of the lowest-energy excited states of α-acylated substrates through time-dependent DFT. The efficiency of the ring contraction at wavelengths longer than those strongly absorbed by the substrates was investigated through wavelength-dependent rate measurements, which revealed a red shift of the photochemical action plot relative to substrate absorbance. The elucidated mechanistic and photophysical details effectively rationalize empirical observations, including additive effects, that were previously poorly understood. Our findings not only demonstrate enhanced synthetic utility of the photomediated ring contraction and shed light on mechanistic details but may also offer valuable guidance for understanding wavelength-dependent reactivity for related photochemical systems.

Identification of Triazolopyrimidinyl Scaffold SARS-CoV-2 Papain-Like Protease (PLpro) Inhibitor

The global impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its companion disease, COVID-19, has reminded us of the importance of basic coronaviral research. In this study, a comprehensive approach using molecular docking, in vitro assays, and molecular dynamics simulations was applied to identify potential inhibitors for SARS-CoV-2 papain-like protease (PLpro), a key and underexplored viral enzyme target. A focused protease inhibitor library was initially created and molecular docking was performed using CmDock software (v0.2.0), resulting in the selection of hit compounds for in vitro testing on the isolated enzyme. Among them, compound 372 exhibited promising inhibitory properties against PLpro, with an IC50 value of 82 ± 34 μM. The compound also displayed a new triazolopyrimidinyl scaffold not yet represented within protease inhibitors. Molecular dynamics simulations demonstrated the favorable binding properties of compound 372. Structural analysis highlighted its key interactions with PLpro, and we stress its potential for further optimization. Moreover, besides compound 372 as a candidate for PLpro inhibitor development, this study elaborates on the PLpro binding site dynamics and provides a valuable contribution for further efforts in pan-coronaviral PLpro inhibitor development.

Molecular medicinal insights into scaffold hopping-based drug discovery success

In both academia and the pharmaceutical industry, innovative hypotheses, methodologies and technologies that can shorten the drug research and development, leading to higher success rates, are vital. In this review, we demonstrate how innovative variations of the scaffold-hopping strategy have been used to create new druggable molecular spaces, drugs, clinical candidates, preclinical candidates, and bioactive agents. We also analyze molecular modulations that enabled improvements of the pharmacodynamic (PD), physiochemical, and pharmacokinetic (PK) properties (P3 properties) of the drugs resulting from these scaffold-hopping strategies.

Alchemical Free Energy Workflows for the Computation of Protein-Ligand Binding Affinities

Alchemical free energy methods can be used for the efficient computation of relative binding free energies during preclinical drug discovery stages. In recent years, this has been facilitated further by the implementation of workflows that enable non-experts to quickly and consistently set up the required simulations. Given the correct input structures, workflows handle the difficult aspects of setting up perturbations, including consistently defining the perturbable molecule, its atom mapping and topology generation, perturbation network generation, running of the simulations via different sampling methods, and analysis of the results. Different academic and commercial workflows are discussed, including FEW, FESetup, FEPrepare, CHARMM-GUI, Transformato, PMX, QLigFEP, TIES, ProFESSA, PyAutoFEP, BioSimSpace, FEP+, Flare, and Orion. These workflows differ in various aspects, such as mapping algorithms or enhanced sampling methods. Some workflows can accommodate more than one molecular dynamics (MD) engine and use external libraries for tasks. Differences between workflows can present advantages for different use cases, however a lack of interoperability of the workflows' components hinders systematic comparisons.

Application of Artificial Intelligence at All Stages of Bone Tissue Engineering

The development of artificial intelligence (AI) has revolutionized medical care in recent years and plays a vital role in a number of areas, such as diagnostics and forecasting. In this review, we discuss the most promising areas of AI application to the field of bone tissue engineering and prosthetics, which can drastically benefit from AI-assisted optimization and patient personalization of implants and scaffolds in ways ranging from visualization and real-time monitoring to the implantation cases prediction, thereby leveraging the compromise between specific architecture decisions, material choice, and synthesis procedure. With the emphasized crucial role of accuracy and robustness of developed AI algorithms, especially in bone tissue engineering, it was shown that rigorous validation and testing, demanding large datasets and extensive clinical trials, are essential, and we discuss how through developing multidisciplinary cooperation among biology, chemistry with materials science, and AI, these challenges can be addressed.

Systematical Evaluation of the Structure-Cardiotoxicity Relationship of 7-Azaindazole-based PI3K Inhibitors Designed by Bioisosteric Approach

A growing concern of cardiotoxicity induced by PI3K inhibitors has raised the requirements to evaluate the structure-cardiotoxicity relationship (SCR) in the development process of novel inhibitors. Based on three bioisosteric 7-azaindazole-based candidate inhibitors namely FD269, FD268 and FD274 that give same order of inhibitory concentration 50% (IC50) magnitude against PI3Ks, in this work, we proposed to systematically evaluate the SCR of 7-azaindazole-based PI3K inhibitors designed by bioisosteric approach. The 24-h lethal concentrations 50% (LC50) of FD269, FD268 and FD274 against zebrafish embryos were 0.35, 4.82 and above 50 μM (not detected), respectively. Determination of the heart rate, pericardial and yolk-sac areas and vascular malformation confirmed the remarkable reduction in the cardiotoxicity of from FD269 to FD268 and to FD274. The IC50s of all three compounds against the hERG channel were tested on the CHO cell line that constitutively expressing hERG channel, which were all higher than 20 μM. The transcriptomic analysis revealed that FD269 and FD268 induced the up-regulation of noxo1b, which encodes a subunit of an NADPH oxidase evoking the oxidative stress. Furthermore, immunohistochemistry tests confirmed the structure-dependent attenuation of the overproduction of ROS and cardiac apoptosis. Our results verified the feasibility of bioisosteric replacement to attenuate the cardiotoxicity of 7-azaindazole-based PI3K inhibitors, suggesting that the screening for PI3K inhibitors with both high potency and low cardiotoxicity from bioisosteres would be a beneficial trial.

Graph Attention Network based mapping of knowledge relations between chemical spaces of Nuclear factor kappa B and Centella asiatica

The confounding nature of the innate immunity target Nuclear Factor kappa B (NF-κB) and its interaction with Centella asiatica (CA) molecules necessitate the intervention of advanced technologies, such as deep learning methods. The integration of chemical space concepts with deep learning technologies is a new way of knowledge mapping used to explore drug-target interactions, especially in molecular libraries derived from traditional medicine based molecular sources. The current constraint of virtual screening for mechanistic target hunting is the use of a binary classification model that includes active and inactive molecules from in vitro experiments to explore drug-target interaction. This study aims to explore the regulatory nature of the molecules from the inhibition and activation of the NF-κB bioassay data set and map this information for a knowledge-based analysis against the molecules of CA, a low-growing tropical plant. This finding has led to a new direction in the field, transitioning from the conventional active-inactive framework to a more comprehensive active-inactive-regulatory model. This approach can be thoroughly explored by leveraging a graph-based deep learning system. The study presents an innovative approach using a Graph Attention Network (GAT) to rank CA molecules in chemical space based on their similarity with NF-κB bioassay molecules, enabling the efficient analysis of complex relationships between molecules and their regulatory function. Graph Attention Network (GAT) overcomes the limitations of traditional deep learning models such as Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) in handling non-Euclidean graph data and allows for a more precise understanding of similarity ranking by utilizing molecular graphs and attention behavior. By measuring similarity and arranging a matrix of similarity ranking based on GAT, deep neural ranking-based algorithms confirmed the regulatory behaviour of an innate immunity target NF-κB with the support of underlying inverse mapping in the surjective chemical spaces of NF-κB bioassays and CA molecular spaces. Overall, the study introduces new techniques for exploring the regulatory behaviour of complex targets like NF-κB. We then used t-SNE for clustering in chemical space and scaffold hunting for scaffold property analysis and identified nine CA molecules that exhibit regulatory behavior of NF-κB target and are recommended for further investigation.

Synthesis and Biological Activity of Silicon-Containing Ethylsulfonylpyridine Insecticides

Ethylsulfonylpyridines are a novel chemical class of insecticides with excellent broad-spectrum activity and an unprecedented mode of action. With the objective of discovering novel ethylsulfonylpyridines with a broader spectrum, wider range of uses, and/or improved properties, we have started a research program aimed at introducing silicon motifs and studying their biological effects. We designed a series of Oxazosulfyl analogues where the hydrogen atom at the 5-position of the pyridyl moiety is replaced by a trialkylsilyl group and prepared these compounds applying denovo synthetic methodology. Our novel ethylsulfonylpyridines exhibit excellent insecticidal activities. The best compound, A18, resulting from our research exhibited an LC50 value of 0.30 mg/L against Plutella xylostella and reached the activity level of the commercial standard Oxazosulfyl. Our findings confirmed our working hypothesis that at the 5-position of the pyridyl moiety larger groups with different hydrophobic, electronic, and steric properties are tolerated.

Alchemical analysis of FDA approved drugs

Chemical space maps help visualize similarities within molecular sets. However, there are many different molecular similarity measures resulting in a confusing number of possible comparisons. To overcome this limitation, we exploit the fact that tools designed for reaction informatics also work for alchemical processes that do not obey Lavoisier's principle, such as the transmutation of lead into gold. We start by using the differential reaction fingerprint (DRFP) to create tree-maps (TMAPs) representing the chemical space of pairs of drugs selected as being similar according to various molecular fingerprints. We then use the Transformer-based RXNMapper model to understand structural relationships between drugs, and its confidence score to distinguish between pairs related by chemically feasible transformations and pairs related by alchemical transmutations. This analysis reveals a diversity of structural similarity relationships that are otherwise difficult to analyze simultaneously. We exemplify this approach by visualizing FDA-approved drugs, EGFR inhibitors, and polymyxin B analogs.

Strategies in the Design and Development of Non-Nucleoside Reverse Transcriptase Inhibitors (NNRTIs)

AIDS (acquired immunodeficiency syndrome) is a potentially life-threatening infectious disease caused by human immunodeficiency virus (HIV). To date, thousands of people have lost their lives annually due to HIV infection, and it continues to be a big public health issue globally. Since the discovery of the first drug, Zidovudine (AZT), a nucleoside reverse transcriptase inhibitor (NRTI), to date, 30 drugs have been approved by the FDA, primarily targeting reverse transcriptase, integrase, and/or protease enzymes. The majority of these drugs target the catalytic and allosteric sites of the HIV enzyme reverse transcriptase. Compared to the NRTI family of drugs, the diverse chemical class of non-nucleoside reverse transcriptase inhibitors (NNRTIs) has special anti-HIV activity with high specificity and low toxicity. However, current clinical usage of NRTI and NNRTI drugs has limited therapeutic value due to their adverse drug reactions and the emergence of multidrug-resistant (MDR) strains. To overcome drug resistance and efficacy issues, combination therapy is widely prescribed for HIV patients. Combination antiretroviral therapy (cART) includes more than one antiretroviral agent targeting two or more enzymes in the life cycle of the virus. Medicinal chemistry researchers apply different optimization strategies including structure- and fragment-based drug design, prodrug approach, scaffold hopping, molecular/fragment hybridization, bioisosterism, high-throughput screening, covalent-binding, targeting highly hydrophobic channel, targeting dual site, and multi-target-directed ligand to identify and develop novel NNRTIs with high antiviral activity against wild-type (WT) and mutant strains. The formulation experts design various delivery systems with single or combination therapies and long-acting regimens of NNRTIs to improve pharmacokinetic profiles and provide sustained therapeutic effects.

Comparative study of the mechanism of natural compounds with similar structures using docking and transcriptome data for improving in silico herbal medicine experimentations

Natural products have successfully treated several diseases using a multi-component, multi-target mechanism. However, a precise mechanism of action (MOA) has not been identified. Systems pharmacology methods have been used to overcome these challenges. However, there is a limitation as those similar mechanisms of similar components cannot be identified. In this study, comparisons of physicochemical descriptors, molecular docking analysis and RNA-seq analysis were performed to compare the MOA of similar compounds and to confirm the changes observed when similar compounds were mixed and used. Various analyses have confirmed that compounds with similar structures share similar MOA. We propose an advanced method for in silico experiments in herbal medicine research based on the results. Our study has three novel findings. First, an advanced network pharmacology research method was suggested by partially presenting a solution to the difficulty in identifying multi-component mechanisms. Second, a new natural product analysis method was proposed using large-scale molecular docking analysis. Finally, various biological data and analysis methods were used, such as in silico system pharmacology, docking analysis and drug response RNA-seq. The results of this study are meaningful in that they suggest an analysis strategy that can improve existing systems pharmacology research analysis methods by showing that natural product-derived compounds with the same scaffold have the same mechanism.

Discovery and Mechanism Study of SARS-CoV-2 3C-like Protease Inhibitors with a New Reactive Group

3CLpro is an attractive target for the treatment of COVID-19. Using the scaffold hopping strategy, we identified a potent inhibitor of 3CLpro (3a) that contains a thiocyanate moiety as a novel warhead that can form a covalent bond with Cys145 of the protein. Tandem mass spectrometry (MS/MS) and X-ray crystallography confirmed the mechanism of covalent formation between 3a and the protein in its catalytic pocket. Moreover, several analogues of compound 3a were designed and synthesized. Among them, compound 3h shows the best inhibition of 3CLpro with an IC50 of 0.322 μM and a kinact/Ki value of 1669.34 M-1 s-1, and it exhibits good target selectivity for 3CLpro against host proteases. Compound 3c inhibits SARS-CoV-2 in Vero E6 cells (EC50 = 2.499 μM) with low cytotoxicity (CC50 > 200 μM). These studies provide ideas and insights to explore and develop new 3CLpro inhibitors in the future.

Azetidines with All-Carbon Quaternary Centers: Merging Relay Catalysis with Strain Release Functionalization

Given the importance and beneficial characteristics of decorated azetidines in medicinal chemistry, efficient strategies for their synthesis are highly sought after. Herein, we report a facile synthesis of the elusive all-carbon quaternary-center-bearing azetidines. By adopting a well-orchestrated polar-radical relay strategy, ring strain release of bench-stable benzoylated 1-azabicyclo[1.1.0]butane (ABB) can be harnessed for nickel-catalyzed Suzuki Csp2-Csp3 cross-coupling with commercially available boronic acids in broad scope (>50 examples), excellent functional group tolerance, and gram-scale utility. Preliminary mechanistic studies provided insights into the underlying mechanism, wherein the ring opening of ABB with a catalytic quantity of bromide accounts for the conversion of ABB into a redox-active azetidine, which subsequently engages in the cross-coupling reaction through a radical pathway. The synergistic bromide and nickel catalysis could intriguingly be derived from a single nickel source (NiBr2). Application of the method to modify natural products, biologically relevant molecules, and pharmaceuticals has been successfully achieved as well as the synthesis of melanocortin-1 receptor (MC-1R) agonist and vesicular acetylcholine transporter (VAChT) inhibitor analogues through bioisosteric replacements of piperidine with azetidine moieties, highlighting the potential of the method in drug optimization studies. Aside from the synthesis of azetidines, we demonstrate the ancillary utility of our nickel catalytic system toward the restricted Suzuki cross-coupling of tertiary alkyl bromides with aryl boronic acids to construct all-carbon quaternary centers.

Oxetane-, Azetidine-, and Bicyclopentane-Bearing N-Heterocycles from Ynones: Scaffold Diversification via Ruthenium-Catalyzed Oxidative Alkynylation

A process for 3-fold scaffold diversification is achieved via ruthenium-catalyzed oxidative alkynylation of commercially available oxetanols, azetidinols and bicyclopentanols to form α,β-acetylenic ketones (ynones), which are subsequently converted to oxetane-, azetidine- and bicyclopentane-bearing pyrazoles, isoxazoles and pyrimidines. A one-pot oxidative alkynylation-condensation protocol that directly converts azetidinols to azetidine-substituted pyrazoles or pyrimidines is demonstrated.

Design, Synthesis, ADME, and Anticancer Studies of Newer N-Aryl-5-(3,4,5-Trifluorophenyl)-1,3,4-Oxadiazol-2-Amines: An Insight into Experimental and Theoretical Investigations

In continuance of our investigation into the anticancer activity of oxadiazoles, we report here the preparation of 10 new 1,3,4-oxadiazole analogues using the scaffold hopping technique. We have prepared the oxadiazoles having a common pharmacophoric structure (oxadiazole linked aryl nucleus) as seen in the reported anticancer agents IMC-038525 (tubulin inhibitor), IMC-094332 (tubulin inhibitor), and FATB (isosteric replacement of the S of thiadiazole with the O of oxadiazole). All of the oxadiazole analogues were predicted for their absorption, distribution, metabolism, and excretion (ADME) profiles and toxicity studies. All of the compounds were found to follow Lipinski's rule of 5 with a safe toxicity profile (Class IV compound) against immunotoxicity, mutagenicity, and toxicity. All of the compounds were synthesized and characterized using spectral data, followed by their anticancer activity tested in a single-dose assay at 10 μM as reported by the National Cancer Institute (NCI US) Protocol against nearly 59 cancer cell lines obtained from nine panels, including non-small-cell lung, ovarian, breast, central nervous system (CNS), colon, leukemia, prostate, and cancer melanoma. N-(2,4-Dimethylphenyl)-5-(3,4,5-trifluorophenyl)-1,3,4-oxadiazol-2-amine (6h) displayed significant anticancer activity against SNB-19, OVCAR-8, and NCI-H40 with percent growth inhibitions (PGIs) of 86.61, 85.26, and 75.99 and moderate anticancer activity against HOP-92, SNB-75, ACHN, NCI/ADR-RES, 786-O, A549/ATCC, HCT-116, MDA-MB-231, and SF-295 with PGIs of 67.55, 65.46, 59.09, 59.02, 57.88, 56.88, 56.53, 56.4, and 51.88, respectively. The compound 6h also registered better anticancer activity than Imatinib against CNS, ovarian, renal, breast, prostate, and melanoma cancers with average PGIs of 56.18, 40.41, 36.36, 27.61, 22.61, and 10.33, respectively. Molecular docking against tubulin, one of the appealing cancer targets, demonstrated an efficient binding within the binding site of combretastatin A4. The ligand 6h (docking score = -8.144 kcal/mol) interacted π-cationically with the residue Lys352 (with the oxadiazole ring). Furthermore, molecular dynamic (MD) simulation studies in complex with the tubulin-combretastatin A4 protein and ligand 6h were performed to examine the dynamic stability and conformational behavior.

Discovery of Novel Hydroxyimine-Tethered Benzenesulfonamides as Potential Human Carbonic Anhydrase IX/XII Inhibitors

To discover novel carbonic anhydrase (CA, EC 4.2.1.1) inhibitors for cancer treatment, a series of 4-{4-[(hydroxyimino)methyl]piperazin-1-yl}benzenesulfonamides were designed and synthesized using SLC-0111 as the lead molecule. The developed novel compounds 27-34 were investigated for the inhibition of human (h) isoforms hCA I, hCA II, hCA IX, and hCA XII. The hCA I was inhibited by compound 29 with a K_i value of 3.0 nM, whereas hCA II was inhibited by compound 32 with a K_i value of 4.4 nM. The tumor-associated hCA IX isoform was inhibited by compound 30 effectively with an K_i value of 43 nM, whereas the activity of another cancer-related isoform, hCA XII, was significantly inhibited by 29 and 31 with a K_i value of 5 nM. Molecular modeling showed that drug molecule 30 participates in significant hydrophobic and hydrogen bond interactions with the active site of the investigated hCAs and binds to zinc through the deprotonated sulfonamide group.

Mode of Action of Novel Pyrazoloquinazoline on Diamondback Moth (Plutella xylostella) Ligand-Gated Chloride Channels

In our previous study, a series of novel pyrazoloquinazolines were synthesized. Pyrazoloquinazoline 5a showed high insecticidal activity against the diamondback moth (Plutella xylostella) and no cross-resistance to fipronil. Patch clamp electrophysiology performed on P. xylostella pupae brains and two-electrode voltage clamp electrophysiology performed on Xenopus Laevis oocytes indicated that 5a might act on the ionotropic γ-aminobutyric acid (GABA) receptor (GABAR) and glutamate-gated chloride channel (GluCl). Moreover, 5a's potency on PxGluCl was about 15-fold higher than on fipronil, which may explain why there was no cross-resistance between 5a and fipronil. Downregulation of the PxGluCl transcription level significantly enhanced the insecticidal activity of 5a on P. xylostella. These findings shed light on the mode of action of 5a and provide important insights into the development of new insecticides for agricultural applications.

Recent developments of imidazo[1,2-a]pyridine analogues as antituberculosis agents

Over the past 2000 years, tuberculosis (TB) has killed more people than any other infectious disease. In 2021, TB claimed 1.6 million lives worldwide, making it the second leading cause of death from an infectious disease after COVID-19. Unfortunately, TB drug discovery research was neglected in the last few decades of the twentieth century. Recently, the World Health Organization has taken the initiative to develop new TB drugs. Imidazopyridine, an important fused bicyclic 5,6 heterocycle has been recognized as a "drug prejudice" scaffold for its wide range of applications in medicinal chemistry. A few examples of imidazo[1,2-a]pyridine exhibit significant activity against multidrug-resistant TB (MDR-TB) and extensively drug-resistant TB (XDR-TB). Here, we critically review anti-TB compounds of the imidazo[1,2-a]pyridine class by discussing their development based on the structure-activity relationship, mode-of-action, and various scaffold hopping strategies over the last decade, which is identified as a renaissance era of TB drug discovery research.

Pharmacological Chaperones and Protein Conformational Diseases: Approaches of Computational Structural Biology

Whenever a protein fails to fold into its native structure, a profound detrimental effect is likely to occur, and a disease is often developed. Protein conformational disorders arise when proteins adopt abnormal conformations due to a pathological gene variant that turns into gain/loss of function or improper localization/degradation. Pharmacological chaperones are small molecules restoring the correct folding of a protein suitable for treating conformational diseases. Small molecules like these bind poorly folded proteins similarly to physiological chaperones, bridging non-covalent interactions (hydrogen bonds, electrostatic interactions, and van der Waals contacts) loosened or lost due to mutations. Pharmacological chaperone development involves, among other things, structural biology investigation of the target protein and its misfolding and refolding. Such research can take advantage of computational methods at many stages. Here, we present an up-to-date review of the computational structural biology tools and approaches regarding protein stability evaluation, binding pocket discovery and druggability, drug repurposing, and virtual ligand screening. The tools are presented as organized in an ideal workflow oriented at pharmacological chaperones' rational design, also with the treatment of rare diseases in mind.

6-Chloro-3-nitro-2-[(phenylsulfonyl)methyl]imidazo[1,2-b]pyridazine

As part of our ongoing scaffold-hopping work on an antikinetoplastid 3-nitroimidazo[1,2-a]pyridine scaffold, we explored 3-nitroimidazo[1,2-b]pyridazine as a potential new antikinetoplastid series. Using conditions previously described by our lab, we obtained 6-chloro-3-nitro-2-[(phenylsulfonyl)methyl]imidazo[1,2-b]pyridazine with 54% yield. In vitro activity of this compound was evaluated against both the promastigote form of Leishmania donovani, the axenic amastigote form of Leishmania infantum and the trypomastigote blood stream form of Trypanosomabrucei brucei, and its influence on cell viability was assessed on the HepG2 cell line. However, despite good activity against the trypomastigote blood stream form of T. b. brucei (EC50 = 0.38 µM), it showed poor solubility in both HepG2 (CC50 > 7.8 µM) and L. infantum axenic amastigotes (EC50 > 1.6 µM) culture media, associated with a loss of activity against the promastigote form of L. infantum (EC50 > 15.6 µM).

Isocyanides in med chem: A scaffold hopping approach for the identification of novel 4-isocyanophenylamides as potent antibacterial agents against methicillin-resistant Staphylococcusaureus

We describe the rational use of the neglected isocyano moiety as pharmacophoric group for the design of novel 4-isocyanophenylamides as antibacterial agents. This class of novel compounds showed to be highly effective against methicillin resistant Staphylococcus aureus strains. In particular, from an extensive screening, we identified compound 42 as lead compound. It has shown a potent antimicrobial activity, an additive effect with most antibiotics currently in use, the ability not to induce the formation of resistant strains after ten passages, and the ability to block the biofilm formation. A nontoxic profile on mammalian cells and a proper metabolic stability on human liver microsome complete the picture of this new weapon against methicillin resistant Staphylococcus aureus infections.

PROTAC technology: A new drug design for chemical biology with many challenges in drug discovery

Target Protein Degradation TPD is a new avenue and revolutionary for therapeutics because redefining the principles of classical drug discovery and guided by event-based target activity rather than the occupancy-driven activity. Since the discovery of the first PROTAC in 2001, TPD represents a rapidly growing technology, with applications in both drug discovery and chemical biology. Over the last decade, many questions have been raised and today the knowledge gained by each team has elucidated a number of them, although there is still a long way to go. The objective of this work is to present the challenges that the PROTAC strategy has very recently addressed in drug design and discovery by presenting extremely recent results from the literature and to provide guidelines in the drug design of new PROTACs as successful therapeutic modality for medicinal chemists.

Indoline-5-Sulfonamides: A Role of the Core in Inhibition of Cancer-Related Carbonic Anhydrases, Antiproliferative Activity and Circumventing of Multidrug Resistance

The overexpression and activity of carbonic anhydrase (CA, EC 4.2.1.1) isoforms CA IX and CA XII promote the accumulation of exceeding protons and acidosis in the extracellular tumor environment. Sulfonamides are effective inhibitors of most families of CAs. In this study, using scaffold-hopping, indoline-5-sulfonamide analogs 4a-u of the CA IX-selective inhibitor 3 were designed and synthesized to evaluate their biological properties. 1-Acylated indoline-5-sulfonamides demonstrated inhibitory activity against tumor-associated CA IX and XII with K_I values up to 132.8 nM and 41.3 nM. Compound 4f, as one of the most potent inhibitors of CA IX and XII, exhibits hypoxic selectivity, suppressing the growth of MCF7 cells at 12.9 µM, and causes partial inhibition of hypoxia-induced CA IX expression in A431 skin cancer cells. 4e and 4f reverse chemoresistance to doxorubicin of K562/4 with overexpression of P-gp.

Scaffold Generator: a Java library implementing molecular scaffold functionalities in the Chemistry Development Kit (CDK)

The concept of molecular scaffolds as defining core structures of organic molecules is utilised in many areas of chemistry and cheminformatics, e.g. drug design, chemical classification, or the analysis of high-throughput screening data. Here, we present Scaffold Generator, a comprehensive open library for the generation, handling, and display of molecular scaffolds, scaffold trees and networks. The new library is based on the Chemistry Development Kit (CDK) and highly customisable through multiple settings, e.g. five different structural framework definitions are available. For display of scaffold hierarchies, the open GraphStream Java library is utilised. Performance snapshots with natural products (NP) from the COCONUT (COlleCtion of Open Natural prodUcTs) database and drug molecules from DrugBank are reported. The generation of a scaffold network from more than 450,000 NP can be achieved within a single day.

Probing Factor Xa Protein-Ligand Interactions: Accurate Free Energy Calculations and Experimental Validations of Two Series of High-Affinity Ligands

The accurate prediction of protein-ligand binding affinity belongs to one of the central goals in computer-based drug design. Molecular dynamics (MD)-based free energy calculations have become increasingly popular in this respect due to their accuracy and solid theoretical basis. Here, we present a combined study which encompasses experimental and computational studies on two series of factor Xa ligands, which enclose a broad chemical space including large modifications of the central scaffold. Using this integrated approach, we identified several new ligands with different heterocyclic scaffolds different from the previously identified indole-2-carboxamides that show superior or similar affinity. Furthermore, the so far underexplored terminal alkyne moiety proved to be a suitable non-classical bioisosteric replacement for the higher halogen-π aryl interactions. With this challenging example, we demonstrated the ability of the MD-based non-equilibrium free energy calculation approach for guiding crucial modifications in the lead optimization process, such as scaffold replacement and single-site modifications at molecular interaction hot spots.

Whole proteome mapping of compound-protein interactions

Off-target binding is one of the primary causes of toxic side effects of drugs in clinical development, resulting in failures of clinical trials. While off-target drug binding is a known phenomenon, experimental identification of the undesired protein binders can be prohibitively expensive due to the large pool of possible biological targets. Here, we propose a new strategy combining chemical similarity principle and deep learning to enable proteome-wide mapping of compound-protein interactions. We have developed a pipeline to identify the targets of bioactive molecules by matching them with chemically similar annotated "bait" compounds and ranking them with deep learning. We have constructed a user-friendly web server for drug-target identification based on chemical similarity (DRIFT) to perform searches across annotated bioactive compound datasets, thus enabling high-throughput, multi-ligand target identification, as well as chemical fragmentation of target-binding moieties.

Drug Chemical Space as a Guide for New Herbicide Development: A Cheminformatic Analysis

Herbicides are critical resources for meeting agricultural demand. While similar in structure and function to pharmaceuticals, the development of new herbicidal mechanisms of action and new scaffolds against known mechanisms of action has been much slower than in pharmaceutical sciences. We hypothesized that this may be due in part to a relative undersampling of possible herbicidal chemistries and set out to test whether this difference in sampling existed and whether increasing the diversity of possible herbicidal chemistries would be likely to result in more efficacious herbicides. To conduct this work, we first identified databases of commercially available herbicides and clinically approved pharmaceuticals. Using these databases, we created a two-dimensional embedding of the chemical, which provides a qualitative visualization of the degree to which each chemotype is distributed within the combined chemical space and shows a moderate degree of overlap between the two sets. Next, we trained several machine learning models to classify herbicides versus drugs based on physicochemical characteristics. The most accurate of these models has an accuracy of 93% with the key differentiating characteristics being the number of polar hydrogens, number of amide bonds, LogP, and polar surface area. We then used several types of scaffold decomposition to quantitatively evaluate the chemical diversity of each molecular family and showed herbicides to have considerably fewer unique structural fragments. Finally, we used molecular docking as an in silico evaluation of further structural diversification in herbicide development. To this end, we identified herbicides with well-characterized binding sites and modified those scaffolds based on similar structural subunits from the drug dataset not present in any commercial herbicide while using the machine-learned model to ensure that required herbicide properties were maintained. Redocking the original and modified scaffolds of several herbicides showed that even this simple design strategy is capable of yielding new molecules with higher predicted affinity for the target enzymes. Overall, we show that herbicides are distinct from drugs based on physicochemical properties but less diverse in their chemistry in a way not governed by these properties. We also demonstrate in silico that increasing the diversity of herbicide scaffolds has the potential to increase potency, potentially reducing the amount needed in agricultural practice.

Discovery of 1'-(1-phenylcyclopropane-carbonyl)-3H-spiro[isobenzofuran-1,3'-pyrrolidin]-3-one as a novel steroid mimetic scaffold for the potent and tissue-specific inhibition of 11β-HSD1 using a scaffold-hopping approach

11β-hydroxysteroid dehydrogenase 1 (11β-HSD1) has been identified as the primary enzyme responsible for the activation of hepatic cortisone to cortisol in specific peripheral tissues resulting in the concomitant antagonism of insulin action within these tissues. Dysregulation of 11β-HSD1, particularly in adipose tissues, has been associated with metabolic syndrome and type 2 diabetes mellitus. Therefore, inhibition of 11β-HSD1 with a small nonsteroidal molecule is therapeutically desirable. Implementation of a scaffold-hopping approach revealed a three-point pharmacophore for 11β-HSD1 that was utilized to design a steroid mimetic scaffold. Reiterative optimization provided valuable insight into the bioactive conformation of our novel scaffold and led to the discovery of INCB13739. Clinical evaluation of INCB13739 confirmed for the first time that tissue-specific inhibition of 11β-HSD1 in patients with type 2 diabetes mellitus was efficacious in controlling glucose levels and reducing cardiovascular risk factors.

Rings in Clinical Trials and Drugs: Present and Future

We present a comprehensive analysis of all ring systems (both heterocyclic and nonheterocyclic) in clinical trial compounds and FDA-approved drugs. We show 67% of small molecules in clinical trials comprise only ring systems found in marketed drugs, which mirrors previously published findings for newly approved drugs. We also show there are approximately 450 000 unique ring systems derived from 2.24 billion molecules currently available in synthesized chemical space, and molecules in clinical trials utilize only 0.1% of this available pool. Moreover, there are fewer ring systems in drugs compared with those in clinical trials, but this is balanced by the drug ring systems being reused more often. Furthermore, systematic changes of up to two atoms on existing drug and clinical trial ring systems give a set of 3902 future clinical trial ring systems, which are predicted to cover approximately 50% of the novel ring systems entering clinical trials.

Local anesthetics via multicomponent reactions

Local anesthetics occupy a prime position in clinical medicine as they temporarily relieve the pain by blocking the voltage-gated sodium channels. However, limited structural diversity, problems with the efficiency of syntheses and increasing toxicity, mean that alternative scaffolds with improved chemical syntheses are urgently needed. Here, we demonstrate an MCR-based approach both towards the synthesis of commercial local anesthetics and towards novel derivatives as potential anesthesia candidates via scaffold hopping. The reactions are efficient and scalable and several single-crystal structures have been obtained. In addition, our methodology has been applied to the synthesis of the antianginal drug ranolazine, via an Ugi three-component reaction. Representative derivatives from our libraries were evaluated as neuronal activity inhibitors using local field potential recordings (LFPs) in mouse hippocampal brain slices and showed very promising results. This study highlights new opportunities in drug discovery targeting local anesthetics.

Exploration of chemical space with partial labeled noisy student self-training and self-supervised graph embedding

BACKGROUND

Drug discovery is time-consuming and costly. Machine learning, especially deep learning, shows great potential in quantitative structure-activity relationship (QSAR) modeling to accelerate drug discovery process and reduce its cost. A big challenge in developing robust and generalizable deep learning models for QSAR is the lack of a large amount of data with high-quality and balanced labels. To address this challenge, we developed a self-training method, Partially LAbeled Noisy Student (PLANS), and a novel self-supervised graph embedding, Graph-Isomorphism-Network Fingerprint (GINFP), for chemical compounds representations with substructure information using unlabeled data. The representations can be used for predicting chemical properties such as binding affinity, toxicity, and others. PLANS-GINFP allows us to exploit millions of unlabeled chemical compounds as well as labeled and partially labeled pharmacological data to improve the generalizability of neural network models.

RESULTS

We evaluated the performance of PLANS-GINFP for predicting Cytochrome P450 (CYP450) binding activity in a CYP450 dataset and chemical toxicity in the Tox21 dataset. The extensive benchmark studies demonstrated that PLANS-GINFP could significantly improve the performance in both cases by a large margin. Both PLANS-based self-training and GINFP-based self-supervised learning contribute to the performance improvement.

CONCLUSION

To better exploit chemical structures as an input for machine learning algorithms, we proposed a self-supervised graph neural network-based embedding method that can encode substructure information. Furthermore, we developed a model agnostic self-training method, PLANS, that can be applied to any deep learning architectures to improve prediction accuracies. PLANS provided a way to better utilize partially labeled and unlabeled data. Comprehensive benchmark studies demonstrated their potentials in predicting drug metabolism and toxicity profiles using sparse, noisy, and imbalanced data. PLANS-GINFP could serve as a general solution to improve the predictive modeling for QSAR modeling.

Visualization of Topological Pharmacophore Space with Graph Edit Distance

A topological pharmacophore (TP) is a chemical graph-based pharmacophore representation, where nodes are pharmacophoric features (PF) and edges are topological distances between PFs. Previously proposed sparse pharmacophore graphs (SPhGs) for TPs were shown to be effective in identifying structurally different active compounds while maintaining the interpretability of the graphs. However, one limitation of using SPhGs as queries is that many structurally similar SPhGs can be identified from a set of active compounds, requiring the classification and visualization of SPhGs, followed by an understanding of the pharmacophore hypotheses. In this study, we propose a scheme for SPhG analysis based on dimensionality reduction techniques with the graph edit distance (GED) metric. This metric enables measuring similarities among SPhGs in a quantitative manner. The visualization of SPhGs, which themselves are the graphs shared by active compounds, can help us understand the pharmacophore hypotheses as well as the data set. As a proof-of-concept study, we generated two-dimensional SPhG-maps using three dimensionality reduction techniques for six biological targets. A comparison with other pharmacophore representations was also conducted. We demonstrated knowledge extraction (interpretation of the data set) from the generated maps. Our findings include a suitable mapping algorithm as well as a pharmacophore hypothesis analysis procedure using an SPhG-map.

Lead generation of cysteine based mesenchymal epithelial transition (c-Met) kinase inhibitors: Using structure-based scaffold hopping, 3D-QSAR pharmacophore modeling, virtual screening, molecular docking, and molecular dynamics simulation

Cysteine-based mesenchymal-epithelial transition (c-Met) is a receptor tyrosine kinase that plays a definitive role during cancer progression and was identified as a possible target for anti-angiogenesis drugs. In the present study, different protocols of computer-based drug design were performed. Construction of predictive pharmacophore model using HypoGen algorithm resulted in a validated model of four features of positive ionizable, hydrogen bond acceptor, hydrophobic, and ring aromatic features with a correlation coefficient of 0.87, a configuration cost of 14.95, and a cost difference of 357.92. The model revealed a promising predictive power and had >90% probability of representing true correlation with the activity data. The model was established using Fisher's validation test at the 95% confidence level and test set prediction (r = 0.96), furthermore, the model was validated by mapping of set of compounds undergoing clinical trials as class Ⅱ c-met inhibitors. The generated valid pharmacophore model was then anticipated for virtual screening of three data bases. Moreover, scaffold hopping using replace fragments protocol was implemented. Hits generated were filtered according to Lipinski's rule; 510 selected hits were anatomized and subjected to molecular docking studies into the crystal structure of c-Met kinase. The good correlation between docking scores and ligand pharmacophore mapping fit values provided a reliable foundation for designing new potentially active candidates that may target c-Met kinase. Eventually, eight hits were selected as potential leads. Subsequently, seven (Hits) have displayed a higher dock score and demonstrated key residue interactions with stable molecular dynamics simulation. Therefore, these c-Met kinase inhibitors may further serve as new chemical spaces in designing new compounds.

FitDock: protein-ligand docking by template fitting

Protein-ligand docking is an essential method in computer-aided drug design and structural bioinformatics. It can be used to identify active compounds and reveal molecular mechanisms of biological processes. A successful docking usually requires thorough conformation sampling and scoring, which are computationally expensive and difficult. Recent studies demonstrated that it can be beneficial to docking with the guidance of existing similar co-crystal structures. In this work, we developed a protein-ligand docking method, named FitDock, which fits initial conformation to the given template using a hierarchical multi-feature alignment approach, subsequently explores the possible conformations and finally outputs refined docking poses. In our comprehensive benchmark tests, FitDock showed 40%-60% improvement in terms of docking success rate and an order of magnitude faster over popular docking methods, if template structures exist (> 0.5 ligand similarity). FitDock has been implemented in a user-friendly program, which could serve as a convenient tool for drug design and molecular mechanism exploration. It is now freely available for academic users at http://cao.labshare.cn/fitdock/.

Application of deep metric learning to molecular graph similarity

Graph based methods are increasingly important in chemistry and drug discovery, with applications ranging from QSAR to molecular generation. Combining graph neural networks and deep metric learning concepts, we expose a framework for quantifying molecular graph similarity based on distance between learned embeddings separate from any endpoint. Using a minimal definition of similarity, and data from the ZINC database of public compounds, this work demonstrate the properties of the embedding and its suitability for a range of applications, among them a novel reconstruction loss method for training deep molecular auto-encoders. Finally, we compare the applications of the embedding to standard practices, with a focus on known failure points and edge cases; concluding that our approach can be used in conjunction to existing methods.

On the Value of Using 3D Shape and Electrostatic Similarities in Deep Generative Methods

Multiparameter optimization, the heart of drug design, is still an open challenge. Thus, improved methods for automated compound design with multiple controlled properties are desired. Here, we present a significant extension to our previously described fragment-based reinforcement learning method (DeepFMPO) for the generation of novel molecules with optimal properties. As before, the generative process outputs optimized molecules similar to the input structures, now with the improved feature of replacing parts of these molecules with fragments of similar three-dimensional (3D) shape and electrostatics. We developed and benchmarked a new python package, ESP-Sim, for the comparison of the electrostatic potential and the molecular shape, allowing the calculation of high-quality partial charges (e.g., RESP with B3LYP/6-31G**) obtained using the quantum chemistry program Psi4. By performing comparisons of 3D fragments, we can simulate 3D properties while overcoming the notoriously difficult step of accurately describing bioactive conformations. The new improved generative (DeepFMPO v3D) method is demonstrated with a scaffold-hopping exercise identifying CDK2 bioisosteres. The code is open-source and freely available.

Free energy perturbation (FEP)-guided scaffold hopping

Scaffold hopping refers to computer-aided screening for active compounds with different structures against the same receptor to enrich privileged scaffolds, which is a topic of high interest in organic and medicinal chemistry. However, most approaches cannot efficiently predict the potency level of candidates after scaffold hopping. Herein, we identified potent PDE5 inhibitors with a novel scaffold via a free energy perturbation (FEP)-guided scaffold-hopping strategy, and FEP shows great advantages to precisely predict the theoretical binding potencies ΔG FEP between ligands and their target, which were more consistent with the experimental binding potencies ΔG EXP (the mean absolute deviations| Δ G FEP - Δ G EXP |  < 2 kcal/mol) than those ΔG MM-PBSA or ΔG MM-GBSA predicted by the MM-PBSA or MM-GBSA method. Lead L12 had an IC50 of 8.7 nmol/L and exhibited a different binding pattern in its crystal structure with PDE5 from the famous starting drug tadalafil. Our work provides the first report via the FEP-guided scaffold hopping strategy for potent inhibitor discovery with a novel scaffold, implying that it will have a variety of future applications in rational molecular design and drug discovery.