Drug Discovery Pipeline Research Articles

Consensus Prediction of Human Gut Microbiota-Mediated Metabolism Susceptibility for Small Molecules by Machine Learning, Structural Alerts, and Dietary Compounds-Based Average Similarity Methods.

The human gut microbiota (HGM) colonizing human gastrointestinal tract (HGT) confers a repertoire of dynamic and unique metabolic capacities that are not possessed by the host and therefore is tentatively perceived as an alternative metabolic ″organ″ besides the liver in the host. Nevertheless, the significant contribution of HGM to the overall human metabolism is often overlooked in the modern drug discovery pipeline. Hence, a systematic evaluation of HGM-mediated drug metabolism is gradually important, and its computational prediction becomes increasingly necessary. In this work, a new data set containing both the HGM-mediated metabolism susceptible (HGMMS) and insusceptible (HGMMI) compounds (329 vs 320) was manually curated. Based on this data set, the first machine learning (ML) model, a new structural alerts (SA) model, and the K-nearest neighboring dietary compounds-based average similarity (AS) model were proposed to directly predict the HGM-mediated metabolism susceptibility for small molecules, and exhibit promising performance on three independent test sets. Finally, consensus prediction (ML/SA/AS) for DrugBank molecules revealed an intriguing phenomenon that a typical Michael acceptor ″α,β-unsaturated carbonyl group″ is a very common warhead for the design of covalent inhibitors and inclined to be metabolized by HGM in anaerobic HGT to generate the reduced metabolite without the reactive warhead, which could be a new concern to medicinal chemists. To the best of our knowledge, we gleaned the first HGMMS/HGMMI data set, developed the first HGMMS/HGMMI classification model, implemented a relatively comprehensive program based on ML/SA/AS approaches, and found a new phenomenon on the HGM-mediated deactivation of an extensively used warhead for covalent inhibitors.

Mechanistic insight into the binding between Ferritin and Serotonin: Possible implications in neurodegenerative diseases

Ferritin is considered to be a crucial iron-binding protein associated with Alzheimer’s disease (AD). The computational approaches, such as molecular docking, molecular dynamics (MD) simulations and principal component analysis (PCA), are helpful to explore potent compounds with predefined targets such as Ferritin in the modern drug discovery pipeline. In this study, the molecular docking calculations revealed that Serotonin has an appreciable binding potential towards Ferritin by forming stable hydrogen bonds (H-bonds). In MD simulation, the analyses of RMSD, RMSF, Rg, SASA, H-bonds, and PCA reveal that the Ferritin-Serotonin docked complex is stable throughout the trajectory. Likewise, the free energy landscape (FEL) analysis exposed that the Ferritin-Serotonin complex stabilized to the global minimum. Overall, the results of this study contribute in numerous aspects to our understanding of Ferritin activity and provide an essential insight into the binding prototype of Serotonin as a potential binding partner of Ferritin to be explored in AD after required experimentation.

Nature-Derived Hit, Lead, and Drug-Like Small Molecules: Current Status and Future Aspects Against Key Target Proteins of Coronaviruses.

COVID-19 pandemic, the most unprecedented event of the year 2020, has brought millions of scientists worldwide in a single platform to fight against it. Though several drugs are now in the clinical trial, few vaccines are available on the market already, but the lack of an effect of those is making the situation worse. In this review, we demonstrated comprehensive data of natural antiviral products showing activities against different proteins of Human Coronaviruses (HCoV) that are responsible for its pathogenesis. Furthermore, we categorized the compounds into the hit, lead, and drug based on the IC50/EC50 value, drug-likeness, and lead-likeness test to portray their potentiality to be a drug. We also demonstrated the present status of our screened antiviral compounds with respect to clinical trials and reported the lead compounds that can be promoted to clinical trial against COVID-19. A systematic search strategy was employed focusing on Natural Products (NPs) with proven activity (in vitro, in vivo, or in silico) against human coronaviruses, in general, and data were gathered from databases like PubMed, Web of Science, Google Scholar, SciVerse, and Scopus. Information regarding clinical trials retrieved from the Clinical Trial Database. Total "245" natural compounds were identified initially from the literature study. Among them, Glycyrrhizin, Caffeic acid, Curcumin is in phase 3, and Tetrandrine, Cyclosporine, Tacrolimus, Everolimus are in phase 4 clinical trial. Except for Glycyrrhizin, all compounds showed activity against COVID-19. In summary, our demonstrated specific small molecules with lead and drug-like capabilities clarified their position in the drug discovery pipeline and proposed future research against COVID-19.

Development of Antibiofilm Therapeutics Strategies to Overcome Antimicrobial Drug Resistance.

A biofilm is a community of stable microorganisms encapsulated in an extracellular matrix produced by themselves. Many types of microorganisms that are found on living hosts or in the environment can form biofilms. These include pathogenic bacteria that can serve as a reservoir for persistent infections, and are culpable for leading to a broad spectrum of chronic illnesses and emergence of antibiotic resistance making them difficult to be treated. The absence of biofilm-targeting antibiotics in the drug discovery pipeline indicates an unmet opportunity for designing new biofilm inhibitors as antimicrobial agents using various strategies and targeting distinct stages of biofilm formation. The strategies available to control biofilm formation include targeting the enzymes and proteins specific to the microorganism and those involved in the adhesion pathways leading to formation of resistant biofilms. This review primarily focuses on the recent strategies and advances responsible for identifying a myriad of antibiofilm agents and their mechanism of biofilm inhibition, including extracellular polymeric substance synthesis inhibitors, adhesion inhibitors, quorum sensing inhibitors, efflux pump inhibitors, and cyclic diguanylate inhibitors. Furthermore, we present the structure–activity relationships (SAR) of these agents, including recently discovered biofilm inhibitors, nature-derived bioactive scaffolds, synthetic small molecules, antimicrobial peptides, bioactive compounds isolated from fungi, non-proteinogenic amino acids and antibiotics. We hope to fuel interest and focus research efforts on the development of agents targeting the uniquely complex, physical and chemical heterogeneous biofilms through a multipronged approach and combinatorial therapeutics for a more effective control and management of biofilms across diseases.

Current Status and Challenges of Human Induced Pluripotent Stem Cell-Derived Liver Models in Drug Discovery.

The pharmaceutical industry is in high need of efficient and relevant in vitro liver models, which can be incorporated in their drug discovery pipelines to identify potential drugs and their toxicity profiles. Current liver models often rely on cancer cell lines or primary cells, which both have major limitations. However, the development of human induced pluripotent stem cells (hiPSCs) has created a new opportunity for liver disease modeling, drug discovery and liver toxicity research. hiPSCs can be differentiated to any cell of interest, which makes them good candidates for disease modeling and drug discovery. Moreover, hiPSCs, unlike primary cells, can be easily genome-edited, allowing the creation of reporter lines or isogenic controls for patient-derived hiPSCs. Unfortunately, even though liver progeny from hiPSCs has characteristics similar to their in vivo counterparts, the differentiation of iPSCs to fully mature progeny remains highly challenging and is a major obstacle for the full exploitation of these models by pharmaceutical industries. In this review, we discuss current liver-cell differentiation protocols and in vitro iPSC-based liver models that could be used for disease modeling and drug discovery. Furthermore, we will discuss the challenges that still need to be overcome to allow for the successful implementation of these models into pharmaceutical drug discovery platforms.

Machine learning-driven identification of drugs inhibiting cytochrome P450 2C9.

Cytochrome P450 2C9 (CYP2C9) is a major drug-metabolizing enzyme that represents 20% of the hepatic CYPs and is responsible for the metabolism of 15% of drugs. A general concern in drug discovery is to avoid the inhibition of CYP leading to toxic drug accumulation and adverse drug–drug interactions. However, the prediction of CYP inhibition remains challenging due to its complexity. We developed an original machine learning approach for the prediction of drug-like molecules inhibiting CYP2C9. We created new predictive models by integrating CYP2C9 protein structure and dynamics knowledge, an original selection of physicochemical properties of CYP2C9 inhibitors, and machine learning modeling. We tested the machine learning models on publicly available data and demonstrated that our models successfully predicted CYP2C9 inhibitors with an accuracy, sensitivity and specificity of approximately 80%. We experimentally validated the developed approach and provided the first identification of the drugs vatalanib, piriqualone, ticagrelor and cloperidone as strong inhibitors of CYP2C9 with IC values <18 μM and sertindole, asapiprant, duvelisib and dasatinib as moderate inhibitors with IC50 values between 40 and 85 μM. Vatalanib was identified as the strongest inhibitor with an IC50 value of 0.067 μM. Metabolism assays allowed the characterization of specific metabolites of abemaciclib, cloperidone, vatalanib and tarafenacin produced by CYP2C9. The obtained results demonstrate that such a strategy could improve the prediction of drug-drug interactions in clinical practice and could be utilized to prioritize drug candidates in drug discovery pipelines.

Special Focus Issue - Targeted protein degradation: a new paradigm in medicinal chemistry.

Special Focus Issue - Targeted protein degradation: a new paradigm in medicinal chemistry.

Pharmacotherapy for Non-Alcoholic Fatty Liver Disease: Emerging Targets and Drug Candidates.

Non-alcoholic fatty liver disease (NAFLD), or metabolic (dysfunction)-associated fatty liver disease (MAFLD), is characterized by high global incidence and prevalence, a tight association with common metabolic comorbidities, and a substantial risk of progression and associated mortality. Despite the increasingly high medical and socioeconomic burden of NAFLD, the lack of approved pharmacotherapy regimens remains an unsolved issue. In this paper, we aimed to provide an update on the rapidly changing therapeutic landscape and highlight the major novel approaches to the treatment of this disease. In addition to describing the biomolecules and pathways identified as upcoming pharmacological targets for NAFLD, we reviewed the current status of drug discovery and development pipeline with a special focus on recent evidence from clinical trials.

Dynamic genetic differentiation drives the widespread structural and functional convergent evolution of snake venom proteinaceous toxins

BackgroundThe explosive radiation and diversification of the advanced snakes (superfamily Colubroidea) was associated with changes in all aspects of the shared venom system. Morphological changes included the partitioning of the mixed ancestral glands into two discrete glands devoted for production of venom or mucous respectively, as well as changes in the location, size and structural elements of the venom-delivering teeth. Evidence also exists for homology among venom gland toxins expressed across the advanced snakes. However, despite the evolutionary novelty of snake venoms, in-depth toxin molecular evolutionary history reconstructions have been mostly limited to those types present in only two front-fanged snake families, Elapidae and Viperidae. To have a broader understanding of toxins shared among extant snakes, here we first sequenced the transcriptomes of eight taxonomically diverse rear-fanged species and four key viperid species and analysed major toxin types shared across the advanced snakes.ResultsTranscriptomes were constructed for the following families and species: Colubridae - Helicops leopardinus, Heterodon nasicus, Rhabdophis subminiatus; Homalopsidae – Homalopsis buccata; Lamprophiidae - Malpolon monspessulanus, Psammophis schokari, Psammophis subtaeniatus, Rhamphiophis oxyrhynchus; and Viperidae – Bitis atropos, Pseudocerastes urarachnoides, Tropidolaeumus subannulatus, Vipera transcaucasiana. These sequences were combined with those from available databases of other species in order to facilitate a robust reconstruction of the molecular evolutionary history of the key toxin classes present in the venom of the last common ancestor of the advanced snakes, and thus present across the full diversity of colubroid snake venoms. In addition to differential rates of evolution in toxin classes between the snake lineages, these analyses revealed multiple instances of previously unknown instances of structural and functional convergences. Structural convergences included: the evolution of new cysteines to form heteromeric complexes, such as within kunitz peptides (the beta-bungarotoxin trait evolving on at least two occasions) and within SVMP enzymes (the P-IIId trait evolving on at least three occasions); and the C-terminal tail evolving on two separate occasions within the C-type natriuretic peptides, to create structural and functional analogues of the ANP/BNP tailed condition. Also shown was that the de novo evolution of new post-translationally liberated toxin families within the natriuretic peptide gene propeptide region occurred on at least five occasions, with novel functions ranging from induction of hypotension to post-synaptic neurotoxicity. Functional convergences included the following: multiple occasions of SVMP neofunctionalised in procoagulant venoms into activators of the clotting factors prothrombin and Factor X; multiple instances in procoagulant venoms where kunitz peptides were neofunctionalised into inhibitors of the clot destroying enzyme plasmin, thereby prolonging the half-life of the clots formed by the clotting activating enzymatic toxins; and multiple occasions of kunitz peptides neofunctionalised into neurotoxins acting on presynaptic targets, including twice just within Bungarus venoms.ConclusionsWe found novel convergences in both structural and functional evolution of snake toxins. These results provide a detailed roadmap for future work to elucidate predator–prey evolutionary arms races, ascertain differential clinical pathologies, as well as documenting rich biodiscovery resources for lead compounds in the drug design and discovery pipeline.

Towards design of drugs and delivery systems with the Martini coarse-grained model.

Coarse-grained (CG) modelling with the Martini force field has come of age. By combining a variety of bead types and sizes with a new mapping approach, the newest version of the model is able to accurately simulate large biomolecular complexes at millisecond timescales. In this perspective, we discuss possible applications of the Martini 3 model in drug discovery and development pipelines and highlight areas for future development. Owing to its high simulation efficiency and extended chemical space, Martini 3 has great potential in the area of drug design and delivery. However, several aspects of the model should be improved before Martini 3 CG simulations can be routinely employed in academic and industrial settings. These include the development of automatic parameterisation protocols for a variety of molecule types, the improvement of backmapping procedures, the description of protein flexibility and the development of methodologies enabling efficient sampling. We illustrate our view with examples on key areas where Martini could give important contributions such as drugs targeting membrane proteins, cryptic pockets and protein-protein interactions and the development of soft drug delivery systems.

Development of a novel machine learning platform to identify structural trends among NNRTI HIV-1 reverse transcriptase inhibitors

With advancements in machine learning powered by on-demand computing and information processing at a large data scale, high throughput virtual screening has become a more attractive method for screening drug candidates, reducing both costs and the timeframe from hit-to-lead. The efficiency of fingerprinting using cheminformatics-based approaches coupled with machine learning to model and identify structure-activity relationships (SAR) has immense potential to improve the time-to-market for the drug development process. This study compared the accuracy of molecular descriptors from two cheminformatics Mordred and PaDEL, software libraries, in characterizing the chemo-structural composition of 53 compounds from the non-nucleoside reverse transcriptase inhibitors (NNRTI) class. We built a logistic regression model to classify NNRTIs based on salient descriptors from each software. We hypothesized that the descriptor data generated by Mordred would be more accurate when characterizing the SAR between NNRTI compounds and the HIV-1 RT enzyme. We identified structural trends in potential inhibitors of the HIV-1 RT enzyme. The classification model built with the filtered set of descriptors from Mordred was superior to the model using PaDEL descriptors as it revealed significant cluster separation between the 53 NNRTI molecules and other drugs, while the filtered PaDEL descriptors model did not show a clear distinction between classes. This approach can accelerate the identification of hit compounds and improve the efficiency of the drug discovery pipeline.

Imaging-Based Machine Learning Analysis of Patient-Derived Tumor Organoid Drug Response

Three-quarters of compounds that enter clinical trials fail to make it to market due to safety or efficacy concerns. This statistic strongly suggests a need for better screening methods that result in improved translatability of compounds during the preclinical testing period. Patient-derived organoids have been touted as a promising 3D preclinical model system to impact the drug discovery pipeline, particularly in oncology. However, assessing drug efficacy in such models poses its own set of challenges, and traditional cell viability readouts fail to leverage some of the advantages that the organoid systems provide. Consequently, phenotypically evaluating complex 3D cell culture models remains difficult due to intra- and inter-patient organoid size differences, cellular heterogeneities, and temporal response dynamics. Here, we present an image-based high-content assay that provides object level information on 3D patient-derived tumor organoids without the need for vital dyes. Leveraging computer vision, we segment and define organoids as independent regions of interest and obtain morphometric and textural information per organoid. By acquiring brightfield images at different timepoints in a robust, non-destructive manner, we can track the dynamic response of individual organoids to various drugs. Furthermore, to simplify the analysis of the resulting large, complex data files, we developed a web-based data visualization tool, the Organoizer, that is available for public use. Our work demonstrates the feasibility and utility of using imaging, computer vision and machine learning to determine the vital status of individual patient-derived organoids without relying upon vital dyes, thus taking advantage of the characteristics offered by this preclinical model system.

Abstract P113: Image-based quantification of immunotherapeutic effect on the tumor-immune interactions in 3D co-cultures

Abstract INTRODUCTION Despite the increasing number of immunotherapies available in preclinical cancer research and drug discovery pipelines, the number of patients benefiting from these immunotherapies remains extremely low. Moreover, the failure of immunotherapies in clinical trials remains very high to date. These high failure rates could be attributed to the lack of deep understanding the mechanism of action of drug candidates and to the absence of clinically relevant preclinical models that are suitable for drug screening purposes. To address these issues, we have developed an in vitro drug screening platform suitable for high-throughput testing in physiologically relevant 3D environment. Our preclinical drug testing platform is based on the co-culture of tumor cells with immune cells in 3D. Image analysis is applied to read out effects of drug candidates on immune cell migration, immune cell infiltration and tumoroid killing. MATERIALS AND METHODS 3D tumor cultures from human cancer cell lines (including breast, lung and colon cancer) or patient-derived colorectal organoids (from HUB Organoid Technology) were generated by embedding tumor cells in 3D ECM-like matrix constituted from protein hydrogel. Subsequently, different immune subsets, including PBMCs from healthy donors, T cells, NK cells or macrophages, were stained with cell tracker and added to the cultures. The cultures were treated with different immune-modulators (e.g. superantigens, activating antibodies, T cell engagers, CSFR1 inhibitor or STING agonist). The effect of immune-modulators on immune cell infiltration and their killing was assessed by high-content imaging and quantified after morphometric analysis with the proprietary Ominer® software. Immune cell-mediated killing was confirmed by measurement of IFNγ secretion in the culture supernatants. RESULTS Image-based analysis allowed for the dissection of complex tumor-immune cell interactions in the 3D cultures. Moreover, morphometric analysis revealed different levels of immune cell infiltration and tumoroid killing upon treatment with different immune-modulators. These effects could be reduced by inhibitory signals from the tumor microenvironment (TME) and were confirmed by the levels of IFNγ secreted by the immune cells. CONCLUSION Our in vitro platform allows quantitative image-based analysis of 3D tumor-immune cell co-cultures in a high-throughput manner, based on spatially resolved information in a more physiologically-relevant setting compared to traditional 2D cultures. This image-based analysis could be employed to dissect the effect of immunotherapy on different cell populations with the TME representing a promising tool to improve our understanding of the mechanism of action of novel treatments, treatment relapse and combination strategies, to eventually lead to a better clinical performance. Citation Format: Lidia Daszkiewicz, Gera Goverse, Nataliia Beztsinna, Marjan van de Merbel, Benjamin Visser, Tomas Veenendaal, Emma Spanjaard, Kuan Yan, Leo Price. Image-based quantification of immunotherapeutic effect on the tumor-immune interactions in 3D co-cultures [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P113.

Deep Learning Algorithms Achieved Satisfactory Predictions When Trained on a Novel Collection of Anticoronavirus Molecules.

Drug discovery and repurposing against COVID-19 is a highly relevant topic with huge efforts dedicated to delivering novel therapeutics targeting SARS-CoV-2. In this context, computer-aided drug discovery is of interest in orienting the early high throughput screenings and in optimizing the hit identification rate. We herein propose a pipeline for Ligand-Based Drug Discovery (LBDD) against SARS-CoV-2. Through an extensive search of the literature and multiple steps of filtering, we integrated information on 2,610 molecules having a validated effect against SARS-CoV and/or SARS-CoV-2. The chemical structures of these molecules were encoded through multiple systems to be readily useful as input to conventional machine learning (ML) algorithms or deep learning (DL) architectures. We assessed the performances of seven ML algorithms and four DL algorithms in achieving molecule classification into two classes: active and inactive. The Random Forests (RF), Graph Convolutional Network (GCN), and Directed Acyclic Graph (DAG) models achieved the best performances. These models were further optimized through hyperparameter tuning and achieved ROC-AUC scores through cross-validation of 85, 83, and 79% for RF, GCN, and DAG models, respectively. An external validation step on the FDA-approved drugs collection revealed a superior potential of DL algorithms to achieve drug repurposing against SARS-CoV-2 based on the dataset herein presented. Namely, GCN and DAG achieved more than 50% of the true positive rate assessed on the confirmed hits of a PubChem bioassay.

Active Learning for Drug Design: A Case Study on the Plasma Exposure of Orally Administered Drugs.

The success of artificial intelligence (AI) models has been limited by the requirement of large amounts of high-quality training data, which is just the opposite of the situation in most drug discovery pipelines. Active learning (AL) is a subfield of AI that focuses on algorithms that select the data they need to improve their models. Here, we propose a two-phase AL pipeline and apply it to the prediction of drug oral plasma exposure. In phase I, the AL-based model demonstrated a remarkable capability to sample informative data from a noisy data set, which used only 30% of the training data to yield a prediction capability with an accuracy of 0.856 on an independent test set. In phase II, the AL-based model explored a large diverse chemical space (855K samples) for experimental testing and feedback. Improved accuracy and new highly confident predictions (50K samples) were observed, which suggest that the model's applicability domain has been significantly expanded.

Pharmacological targeting of Sam68 functions in colorectal cancer stem cells

SummaryCancer stem cells (CSCs) are documented to play a key role in tumorigenesis and therapy resistance. Despite significant progress in clinical oncology, CSC reservoirs remain elusive and difficult to eliminate. Reverse-turn peptidomimetics were characterized as disruptors of CBP/beta-Catenin interactions and represent a promising avenue to curb hyperactive canonical Wnt/beta-Catenin signaling in CSCs. Recent studies suggested Sam68 as a critical mediator of reverse-turn peptidomimetics response in CSC populations. Using computational and biochemical approaches we confirmed Sam68 as a primary target of reverse-turn peptidomimetics. Furthermore, we executed an in silico drug discovery pipeline to identify yet uncharacterized reverse-turn peptidomimetic structures displaying superior anti-CSC activity in transformed pluripotent and colorectal cancer cell models. Thus, we identified YB-0158 as a reverse-turn peptidomimetic small molecule with enhanced translational potential, altering key hallmarks of human colorectal CSCs in patient-derived ex vivo organoids and in vivo serial tumor transplantation.

Oral Efficacy of a Diselenide Compound Loaded in Nanostructured Lipid Carriers in a Murine Model of Visceral Leishmaniasis.

Leishmaniasis urgently needs new oral treatments, as it is one of the most important neglected tropical diseases that affects people with poor resources. The drug discovery pipeline for oral administration currently discards entities with poor aqueous solubility and permeability (class IV compounds in the Biopharmaceutical Classification System, BCS) such as the diselenide 2m, a trypanothione reductase (TR) inhibitor. This work was assisted by glyceryl palmitostearate and diethylene glycol monoethyl ether-based nanostructured lipid carriers (NLC) to render 2m bioavailable and effective after its oral administration. The loading of 2m in NLC drastically enhanced its intestinal permeability and provided plasmatic levels higher than its effective concentration (IC50). In L. infantum-infected BALB/c mice, 2m-NLC reduced the parasite burden in the spleen, liver, and bone marrow by at least 95% after 5 doses, demonstrating similar efficacy as intravenous Fungizone. Overall, compound 2m and its formulation merit further investigation as an oral treatment for visceral leishmaniasis.

Mechanistic Investigation of Drug Supersaturation in the Presence of Polysorbates as Solubilizing Additives by Solution Nuclear Magnetic Resonance Spectroscopy.

The introduction of solubilizing additives has historically been an attractive approach to address the ever-growing proportion of poorly water-soluble drug (PWSD) compounds within the modern drug discovery pipeline. Lipid-formulations, and more specifically micelle formulations, have garnered particular interest because of their simplicity, size, scalability, and avoidance of solid-state limitations. Although micelle formulations have been widely utilized, the molecular mechanism of drug solubilization in surfactant micelles is still poorly understood. In this study, a series of modern nuclear magnetic resonance (NMR) methods are utilized to gain a molecular-level understanding of intermolecular interactions and kinetics in a model system. This approach enabled the understanding of how a PWSD, 17β-Estradiol (E2), solubilizes within a nonionic micelle system composed of polysorbate 80 (PS80). Based on one-dimensional (1D) 1H chemical shift differences of E2 in PS80 solutions, as well as intermolecular correlations established from 1D selective nuclear Overhauser effect (NOE) and two-dimensional NOE spectroscopy experiments, E2 was found to accumulate within the palisade layer of PS80 micelles. A potential hydrogen-bonding interaction between a hydroxyl group of E2 and a carbonyl group of PS80 alkane chains may allow for stabilizing E2-PS80 mixed micelles. Diffusion and relaxation NMR analysis and particle size measurements using dynamic light scattering indicate a slight increase in the micellar size with increasing degrees of supersaturation, resulting in slower mobility of the drug molecule. Based on these structural findings, a theoretical orientation model of E2 molecules with PS80 molecules was developed and validated by computational docking simulations.

Drug Therapeutic-Use Class Prediction and Repurposing Using Graph Convolutional Networks.

An important stage in the process of discovering new drugs is when candidate molecules are tested of their efficacy. It is reported that testing drug efficacy empirically costs billions of dollars in the drug discovery pipeline. As a mechanism of expediting this process, researchers have resorted to using computational methods to predict the action of molecules in silico. Here, we present a way of predicting the therapeutic-use class of drugs from chemical structures only using graph convolutional networks. In comparison with existing methods which use fingerprints or images as training samples, our approach has yielded better results in all metrics under consideration. In particular, validation accuracy increased from 83–88% to 86–90% for single label tasks. Similarly, the model achieved an accuracy of over 88% on new test data. Finally, our multi-label classification model made new predictions which indicated that some of the drugs could have other therapeutic uses other than those indicated in the dataset. We performed a literature-based evaluation of these predictions and found evidence that validates them. This renders the model a potential tool to be used in search of drugs that are candidates for repurposing.

Prediction of the chromatographic hydrophobicity index with immobilized artificial membrane chromatography using simple molecular descriptors and artificial neural networks

Screening of physicochemical properties should be considered one of the essential steps in the drug discovery pipeline. Among the available methods, biomimetic chromatography with an immobilized artificial membrane is a powerful tool for simulating interactions between a molecule and a biological membrane. This study developed a quantitative structure–retention relationships model that would predict the chromatographically determined affinity of xenobiotics to phospholipids, expressed as a chromatographic hydrophobicity index determined using immobilized artificial membrane chromatography. A heterogeneous set of 261 molecules, mostly showing pharmacological activity or toxicity, was analyzed chromatographically to realize this goal. The chromatographic analysis was performed using the fast gradient protocol proposed by Valko, where acetonitrile was applied as an organic modifier. Next, quantitative structure–retention relationships modeling was performed using multiple linear regression (MLR) methods and artificial neural networks (ANNs) coupled with genetic algorithm (GA)-inspired selection. Subsequently, the selection of the best ANN was supported by statistical parameters, the sum of ranking differences approach with the comparison of rank by random numbers and hierarchical cluster analysis.