Early Stages Of Drug Discovery Research Articles

Machine Learning for In Silico ADMET Prediction.

ADMET (absorption, distribution, metabolism, excretion, and toxicity) describes a drug molecule's pharmacokinetics and pharmacodynamics properties. ADMET profile of a bioactive compound can impact its efficacy and safety. Moreover, efficacy and safety are considered some of the major causes of clinical attrition in the development of new chemical entities. In past decades, various machine learning or quantitative structure-activity relationship (QSAR) methods have been successfully integrated in the modeling of ADMET. Recent advances have been made in the collection of data and the development of various in silico methods to assess and predict ADMET of bioactive compounds in the early stages of drug discovery and development process.

Network-Driven Drug Discovery.

We describe an approach to early stage drug discovery that explicitly engages with the complexities of human biology. The combined computational and experimental approach is formulated on a conceptual framework in which network biology is used to bridge between individual molecular entities and the cellular phenotype that emerges when those entities interact in a network. Multiple aspects of early stage discovery are addressed including the data-driven elucidation of biological processes implicated in disease, target identification and validation, phenotypic discovery of active molecules and their mechanism of action, and extraction of genetic target support from human population genetics data. Validation is described via summary of a number of discovery projects and details from a project aimed at COVID-19 disease.

Validation and invalidation of SARS-CoV-2 main protease inhibitors using the Flip-GFP and Protease-Glo luciferase assays

SARS-CoV-2 main protease (Mpro) is one of the most extensively exploited drug targets for COVID-19. Structurally disparate compounds have been reported as Mpro inhibitors, raising the question of their target specificity. To elucidate the target specificity and the cellular target engagement of the claimed Mpro inhibitors, we systematically characterize their mechanism of action using the cell-free FRET assay, the thermal shift-binding assay, the cell lysate Protease-Glo luciferase assay, and the cell-based FlipGFP assay. Collectively, our results have shown that majority of the Mpro inhibitors identified from drug repurposing including ebselen, carmofur, disulfiram, and shikonin are promiscuous cysteine inhibitors that are not specific to Mpro, while chloroquine, oxytetracycline, montelukast, candesartan, and dipyridamole do not inhibit Mpro in any of the assays tested. Overall, our study highlights the need of stringent hit validation at the early stage of drug discovery.

Development of a Fluorescent-Labeled Trapping Reagent to Detect Reactive Acyl Glucuronides.

Acyl glucuronides are common metabolites of carboxylic acid-containing compounds. Since acyl glucuronides sometimes show high reactivity, they are considered to be involved in drug toxicity. Therefore, it is important to evaluate the risk posed by acyl glucuronides in the development of safe drugs; however, there are no suitable evaluation methods for the early stages of drug discovery. We aimed to develop a trapping reagent that detects reactive acyl glucuronides to assess their risk. We designed a diamine-structured trapping reagent, Dap-Dan, and compared its trapping ability with the reported one that has an amino group, and results showed that Dap-Dan showed higher accuracy. In the trapping assay with 17 medicines containing a carboxylic acid, Dap-Dan trapped acyl glucuronides that had a higher risk of toxicity. In conclusion, Dap-Dan can be useful for evaluating the risk of reactive acyl glucuronides.

Evaluation of FOXO1 Target Engagement Using a Single-Cell Microfluidic Platform.

The cellular thermal shift assay (CETSA) has been used extensively since its introduction to study drug-target engagement within both live cells and cellular lysate. This has proven to be a useful tool in early stage drug discovery and is used to study a wide range of protein classes. We describe the application of a single-cell CETSA workflow within a microfluidic affinity capture (MAC) chip. This has enabled us to quantitatively determine the active FOXO1 single-molecule count and observe FOXO1 stabilization and destabilization in the presence of three small molecule inhibitors, including demonstrating the determination of EC50. The successful use of the MAC chip for single-cell CETSA paves the way for the study of precious clinical samples owing to the low number of cells needed by the chip. It also provides a useful tool for studying any underlying population heterogeneity that exists within a cellular system, a feature that is usually masked when conducting ensemble measurements.

Quantitative Estimate Index for Early-Stage Screening of Compounds Targeting Protein-Protein Interactions.

Drug-likeness quantification is useful for screening drug candidates. Quantitative estimates of drug-likeness (QED) are commonly used to assess quantitative drug efficacy but are not suitable for screening compounds targeting protein-protein interactions (PPIs), which have recently gained attention. Therefore, we developed a quantitative estimate index for compounds targeting PPIs (QEPPI), specifically for early-stage screening of PPI-targeting compounds. QEPPI is an extension of the QED method for PPI-targeting drugs that models physicochemical properties based on the information available for drugs/compounds, specifically those reported to act on PPIs. FDA-approved drugs and compounds in iPPI-DB, which comprise PPI inhibitors and stabilizers, were evaluated using QEPPI. The results showed that QEPPI is more suitable than QED for early screening of PPI-targeting compounds. QEPPI was also considered an extended concept of the “Rule-of-Four” (RO4), a PPI inhibitor index. We evaluated the discriminatory performance of QEPPI and RO4 for datasets of PPI-target compounds and FDA-approved drugs using F-score and other indices. The F-scores of RO4 and QEPPI were 0.451 and 0.501, respectively. QEPPI showed better performance and enabled quantification of drug-likeness for early-stage PPI drug discovery. Hence, it can be used as an initial filter to efficiently screen PPI-targeting compounds.

The Discovery of Peptide Macrocycle Rescuers of Pathogenic Protein Misfolding and Aggregation by Integrating SICLOPPS Technology and Ultrahigh-Throughput Screening in Bacteria.

The phenomenon of protein misfolding and aggregation has been widely associated with numerous human diseases, such as Alzheimer's disease, systemic amyloidosis and type 2 diabetes, the vast majority of which remain incurable. To advance early stage drug discovery against these diseases, investigation of molecular libraries with expanded diversities and ultrahigh-throughput screening methodologies that allow deeper investigation of chemical space are urgently required. Toward this, we describe how Escherichia coli can be engineered so as to enable (1) the production of expanded combinatorial libraries of short, drug-like, head-to-tail cyclic peptides and (2) their simultaneous functional screening for identifying effective inhibitors of protein misfolding and aggregation using a genetic assay that links protein folding and misfolding to cell fluorescence. In this manner, cyclic peptides with the ability to inhibit pathogenic protein misfolding and/or aggregation can be readily selected by flow cytometric cell sorting in an ultrahigh-throughput fashion. This biotechnological approach accelerates significantly the identification of hit/lead molecules with potentially therapeutic properties against devastating diseases.

Investigating receptor-mediated antibody transcytosis using blood\u2013brain barrier organoid arrays

BackgroundThe pathways that control protein transport across the blood–brain barrier (BBB) remain poorly characterized. Despite great advances in recapitulating the human BBB in vitro, current models are not suitable for systematic analysis of the molecular mechanisms of antibody transport. The gaps in our mechanistic understanding of antibody transcytosis hinder new therapeutic delivery strategy development.MethodsWe applied a novel bioengineering approach to generate human BBB organoids by the self-assembly of astrocytes, pericytes and brain endothelial cells with unprecedented throughput and reproducibility using micro patterned hydrogels. We designed a semi-automated and scalable imaging assay to measure receptor-mediated transcytosis of antibodies. Finally, we developed a workflow to use CRISPR/Cas9 gene editing in BBB organoid arrays to knock out regulators of endocytosis specifically in brain endothelial cells in order to dissect the molecular mechanisms of receptor-mediated transcytosis.ResultsBBB organoid arrays allowed the simultaneous growth of more than 3000 homogenous organoids per individual experiment in a highly reproducible manner. BBB organoid arrays showed low permeability to macromolecules and prevented transport of human non-targeting antibodies. In contrast, a monovalent antibody targeting the human transferrin receptor underwent dose- and time-dependent transcytosis in organoids. Using CRISPR/Cas9 gene editing in BBB organoid arrays, we showed that clathrin, but not caveolin, is required for transferrin receptor-dependent transcytosis.ConclusionsHuman BBB organoid arrays are a robust high-throughput platform that can be used to discover new mechanisms of receptor-mediated antibody transcytosis. The implementation of this platform during early stages of drug discovery can accelerate the development of new brain delivery technologies.

PredMS: a random forest model for predicting metabolic stability of drug candidates in human liver microsomes.

Poor metabolic stability leads to drug development failure. Therefore, it is essential to evaluate the metabolic stability of small compounds for successful drug discovery and development. However, evaluating metabolic stability in vitro and in vivo is expensive, time-consuming and laborious. In addition, only a few free software programs are available for metabolic stability data and prediction. Therefore, in this study, we aimed to develop a prediction model that predicts the metabolic stability of small compounds. We developed a computational model, PredMS, which predicts the metabolic stability of small compounds as stable or unstable in human liver microsomes. PredMS is based on a random forest model using an in-house database of metabolic stability data of 1917 compounds. To validate the prediction performance of PredMS, we generated external test data of 61 compounds. PredMS achieved an accuracy of 0.74, Matthew's correlation coefficient of 0.48, sensitivity of 0.70, specificity of 0.86, positive predictive value of 0.94 and negative predictive value of 0.46 on the external test dataset. PredMS will be a useful tool to predict the metabolic stability of small compounds in the early stages of drug discovery and development. The source code for PredMS is available at https://bitbucket.org/krictai/predms, and the PredMS web server is available at https://predms.netlify.app. Supplementary data are available at Bioinformatics online.

Small footprint liquid chromatography-mass spectrometry for pharmaceutical reaction monitoring and automated process analysis

Liquid chromatography (LC) has broad applicability in the pharmaceutical industry, from the early stages of drug discovery to reaction monitoring and process control. However, small footprint, truly portable LC systems have not yet been demonstrated and fully evaluated practically for on-line, in-line or at-line pharmaceutical analysis. Herein, a portable, briefcase-sized capillary LC fitted with a miniature multi-deep UV-LED detector has been developed and interfaced with a portable mass spectrometer for on-site pharmaceutical analysis. With this configuration, the combined small footprint portable LC-UV/MS system was utilized for the determination of small molecule pharmaceuticals and reaction monitoring. The LC-UV/MS system was interfaced directly with a process sample cart and applied to automated pharmaceutical analysis, as well as also being benchmarked against a commercial process UPLC system (Waters PATROL system). The portable system gave low detection limits (∼3 ppb), a wide dynamic range (up to 200 ppm) and was used to confirm the identity of reaction impurities and for studying the kinetics of synthesis. The developed platform showed robust performance for automated process analysis, with less than 5.0% relative standard deviation (RSD) on sample-to-sample reproducibility, and less than 2% carryover between samples. The system has been shown to significantly increase throughput by providing near real-time analysis and to improve understanding of synthetic processes.

Contact Map Fingerprints of Protein-Ligand Unbinding Trajectories Reveal Mechanisms Determining Residence Times Computed from Scaled Molecular Dynamics.

The binding kinetic properties of potential drugs may significantly influence their subsequent clinical efficacy. Predictions of these properties based on computer simulations provide a useful alternative to their expensive and time-consuming experimental counterparts, even at an early drug discovery stage. Herein, we perform scaled molecular dynamics (ScaledMD) simulations on a set of 27 ligands of HSP90 belonging to more than seven chemical series to estimate their relative residence times. We introduce two new techniques for the analysis and the classification of the simulated unbinding trajectories. The first technique, which helps in estimating the limits of the free energy well around the bound state, and the second one, based on a new contact map fingerprint, allow the description and the comparison of the paths that lead to unbinding. Using these analyses, we find that ScaledMD's relative residence time generally enables the identification of the slowest unbinders. We propose an explanation for the underestimation of the residence times of a subset of compounds, and we investigate how the biasing in ScaledMD can affect the mechanistic insights that can be gained from the simulations.

SimilarityLab: Molecular Similarity for SAR Exploration and Target Prediction on the Web

Exploration of chemical space around hit, experimental, and known active compounds is an important step in the early stages of drug discovery. In academia, where access to chemical synthesis efforts is restricted in comparison to the pharma-industry, hits from primary screens are typically followed up through purchase and testing of similar compounds, before further funding is sought to begin medicinal chemistry efforts. Rapid exploration of druglike similars and structure–activity relationship profiles can be achieved through our new webservice SimilarityLab. In addition to searching for commercially available molecules similar to a query compound, SimilarityLab also enables the search of compounds with recorded activities, generating consensus counts of activities, which enables target and off-target prediction. In contrast to other online offerings utilizing the USRCAT similarity measure, SimilarityLab’s set of commercially available small molecules is consistently updated, currently containing over 12.7 million unique small molecules, and not relying on published databases which may be many years out of date. This ensures researchers have access to up-to-date chemistries and synthetic processes enabling greater diversity and access to a wider area of commercial chemical space. All source code is available in the SimilarityLab source repository.

In vivo-in vitro correlation of antitumor activity of heat shock protein 90 (HSP90) inhibitors with a pharmacokinetics/pharmacodynamics analysis using NCI-N87 xenograft mice

The in vitro antitumor activity (e.g. IC50) of anticancer drugs is important for selecting candidate compounds for in vivo drug efficacy study in the early stage of drug discovery. In this study, we investigated the relationship between in vitro IC50 and in vivo EC50 using six heat shock protein 90 (HSP90) inhibitors. IC50 of each compound was calculated from in vitro cell proliferation assays using the NCI-N87 cancer cell line. Each compound was administered to NCI-N87 xenograft mice, and EC50 and the maximum tumour-killing rate constant were calculated from pharmacokinetics/pharmacodynamics analyses using plasma concentrations and tumour volumes. IC50 obtained in vitro was poorly correlated with EC50 obtained in vivo, while a good correlation (r = 0.856) was observed between them when corrected with the unbound fraction ratio. The results of this study using of HSP90 inhibitors as model compounds suggest importance of the consideration of an unbound fraction to evaluate the relationship between IC50 and EC50. These results will contribute to improvement in the prediction accuracy of in vivo drug efficacy from in vitro activity and the efficiency of drug discovery research.

Interpretation of Ligand-Based Activity Cliff Prediction Models Using the Matched Molecular Pair Kernel.

Activity cliffs (ACs) are formed by two structurally similar compounds with a large difference in potency. Accurate AC prediction is expected to help researchers’ decisions in the early stages of drug discovery. Previously, predictive models based on matched molecular pair (MMP) cliffs have been proposed. However, the proposed methods face a challenge of interpretability due to the black-box character of the predictive models. In this study, we developed interpretable MMP fingerprints and modified a model-specific interpretation approach for models based on a support vector machine (SVM) and MMP kernel. We compared important features highlighted by this SVM-based interpretation approach and the SHapley Additive exPlanations (SHAP) as a major model-independent approach. The model-specific approach could capture the difference between AC and non-AC, while SHAP assigned high weights to the features not present in the test instances. For specific MMPs, the feature weights mapped by the SVM-based interpretation method were in agreement with the previously confirmed binding knowledge from X-ray co-crystal structures, indicating that this method is able to interpret the AC prediction model in a chemically intuitive manner.

Computational prediction of frequent hitters in target-based and cell-based assays

Compounds interfering with high-throughput screening (HTS) assay technologies (also known as “badly behaving compounds”, “bad actors”, “nuisance compounds” or “PAINS”) pose a major challenge to early-stage drug discovery. Many of these problematic compounds are “frequent hitters”, and we have recently published a set of machine learning models (“Hit Dexter 2.0”) for flagging such compounds.Here we present a new generation of machine learning models which are derived from a large, manually curated and annotated data set. For the first time, these models cover, in addition to target-based assays, also cell-based assays. Our experiments show that cell-based assays behave indeed differently from target-based assays, with respect to hit rates and frequent hitters, and that dedicated models are required to produce meaningful predictions. In addition to these extensions and refinements, we explored a variety of additional setups for modeling, including the combination of four machine learning classifiers (i.e. k-nearest neighbors (KNN), extra trees, random forest and multilayer perceptron) with four sets of descriptors (Morgan2 fingerprints, Morgan3 fingerprints, MACCS keys and 2D physicochemical property descriptors).Testing on holdout data as well as data sets of “dark chemical matter” (i.e. compounds that have been extensively tested in biological assays but have never shown activity) and known bad actors show that the multilayer perceptron classifiers in combination with Morgan2 fingerprints outperform other setups in most cases. The best multilayer perceptron classifiers obtained Matthews correlation coefficients of up to 0.648 on holdout data. These models are available via a free web service.

Deep learning detects cardiotoxicity in a high-content screen with induced pluripotent stem cell-derived cardiomyocytes

Drug-induced cardiotoxicity and hepatotoxicity are major causes of drug attrition. To decrease late-stage drug attrition, pharmaceutical and biotechnology industries need to establish biologically relevant models that use phenotypic screening to detect drug-induced toxicity in vitro. In this study, we sought to rapidly detect patterns of cardiotoxicity using high-content image analysis with deep learning and induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs). We screened a library of 1280 bioactive compounds and identified those with potential cardiotoxic liabilities in iPSC-CMs using a single-parameter score based on deep learning. Compounds demonstrating cardiotoxicity in iPSC-CMs included DNA intercalators, ion channel blockers, epidermal growth factor receptor, cyclin-dependent kinase, and multi-kinase inhibitors. We also screened a diverse library of molecules with unknown targets and identified chemical frameworks that show cardiotoxic signal in iPSC-CMs. By using this screening approach during target discovery and lead optimization, we can de-risk early-stage drug discovery. We show that the broad applicability of combining deep learning with iPSC technology is an effective way to interrogate cellular phenotypes and identify drugs that may protect against diseased phenotypes and deleterious mutations.

Luminescence Energy Transfer-Based Screening and Target Engagement Approaches for Chemical Biology and Drug Discovery.

Luminescence is characterized by the spontaneous emission of light resulting from either chemical or biological reactions. Because of their high sensitivity, reduced background interference, and applicability to numerous situations, luminescence-based assay strategies play an essential role in early-stage drug discovery. Newer developments in luminescence-based technologies have dramatically affected the ability of researchers to investigate molecular binding events. At the forefront of these developments are the nano bioluminescence resonance energy transfer (NanoBRET) and amplified luminescent proximity homogeneous assay (Alpha) technologies. These technologies have opened up numerous possibilities for analyzing the molecular biophysical properties of complexes in environments such as cell lysates. Moreover, NanoBRET enables the validation and quantitation of the interactions between therapeutic targets and small molecules in live cells, representing an essential benchmark for preclinical drug discovery. Both techniques involve proximity-based luminescence energy transfer, in which excited-state energy is transferred from a donor to an acceptor, where the efficiency of transfer depends on proximity. Both approaches can be applied to high-throughput compound screening in biological samples, with the NanoBRET assay providing opportunities for live-cell screening. Representative applications of both technologies for assessing physical interactions and associated challenges are discussed.

ADME prediction with KNIME: A retrospective contribution to the second "Solubility Challenge".

Computational models for predicting aqueous solubility from the molecular structure represent a promising strategy from the perspective of drug design and discovery. Since the first “Solubility Challenge”, these initiatives have marked the state-of-art of the modelling algorithms used to predict drug solubility. In this regard, the quality of the input experimental data and its influence on model performance has been frequently discussed. In our previous study, we developed a computational model for aqueous solubility based on recursive random forest approaches. The aim of the current commentary is to analyse the performance of this already trained predictive model on the molecules of the second “Solubility Challenge”. Even when our training set has inconsistencies related to the pH, solid form and temperature conditions of the solubility measurements, the model was able to predict the two sets from the second “Solubility Challenge” with statistics comparable to those of the top ranked models. Finally, we provided a KNIME automated workflow to predict aqueous solubility of new drug candidates, during the early stages of drug discovery and development, for ensuring the applicability and reproducibility of our model.

Miniaturization and Automation of a Human In Vitro Blood-Brain Barrier Model for the High-Throughput Screening of Compounds in the Early Stage of Drug Discovery.

Central nervous system (CNS) diseases are one of the top causes of death worldwide. As there is a difficulty of drug penetration into the brain due to the blood–brain barrier (BBB), many CNS drugs treatments fail in clinical trials. Hence, there is a need to develop effective CNS drugs following strategies for delivery to the brain by better selecting them as early as possible during the drug discovery process. The use of in vitro BBB models has proved useful to evaluate the impact of drugs/compounds toxicity, BBB permeation rates and molecular transport mechanisms within the brain cells in academic research and early-stage drug discovery. However, these studies that require biological material (animal brain or human cells) are time-consuming and involve costly amounts of materials and plastic wastes due to the format of the models. Hence, to adapt to the high yields needed in early-stage drug discoveries for compound screenings, a patented well-established human in vitro BBB model was miniaturized and automated into a 96-well format. This replicate met all the BBB model reliability criteria to get predictive results, allowing a significant reduction in biological materials, waste and a higher screening capacity for being extensively used during early-stage drug discovery studies.

Toxicity of Jegosaponins A and B from Styrax japonica Siebold et al. Zuccarini in Prostate Cancer Cells and Zebrafish Embryos Resulting from Increased Membrane Permeability.

(1) Background: Screening of medicinal herbs is one of the most powerful approaches to identifying novel therapeutic molecules against many human diseases. To avoid potential harmful effects during medicinal use, toxicity testing is necessary in the early stages of drug discovery. The objective of this study was to identify the cytotoxic mechanisms of jegosaponin A and B from Styrax japonica Siebold et al. Zuccarini; (2) Methods: We screened Japanese medicinal herb extracts using PC-3 prostate cancer cells and found that a methanol extract isolated from the unripe fruit of Styrax japonica Siebold et al. Zuccarini (SJSZ) had an inhibitory effect on cell viability. We further performed fractionation assays with PC-3 cells and identified the bioactive compounds using LC/MS and NMR analysis. We clarified the toxic mechanisms of these compounds using PC-3 cells and zebrafish embryos; (3) Results: We identified two active molecules, jegosaponin A and jegosaponin B, in the inhibitory fractions of the methanol extract. These jegosaponins are toxic to zebrafish embryos during the early developmental stage. Jegosaponin A and B showed strong haemolytic activity in sheep defibrinated blood (EC50 = 2.1 μM, and 20.2 μM, respectively) and increased the cell membrane permeability in PC-3 cells and zebrafish embryos, which were identified using a membrane non-permeable DRAQ7, a fluorescent nucleus staining dye; (4) We identified the cytotoxic compounds jegosaponin A and B from SJSZ, which we showed to exhibit cell membrane disruptive properties using cell- and zebrafish-based testing.