Early-stage Drug Research Articles

Ringtail: A Python Tool for Efficient Management and Storage of Virtual Screening Results.

Virtual screening using molecular docking is now routinely used for the rapid evaluation of very large ligand libraries in early stage drug discovery. As the size of compound libraries which can feasibly be screened grows, so do the challenges in result management and storage. Here we introduce Ringtail, a new Python tool in the AutoDock Suite for efficient storage and analysis of virtual screening data based on portable SQLite databases. Ringtail is designed to work with AutoDock-GPU and AutoDock Vina out-of-the-box. Its modular design also allows for easy extension to support input file types from other docking software, different storage solutions, and incorporation into other applications. Ringtail's SQLite database output can dramatically reduce the required disk storage (36-46 fold) by selecting individual poses to store and by taking advantage of the relational database format. Filtering times are also dramatically reduced, requiring minutes to filter millions of ligands. Thus, Ringtail is a tool that can immediately integrate into existing virtual screening pipelines using AutoDock-GPU and Vina, and is scriptable and modifiable to fit specific user needs.

Digital Research Environment(DRE)-enabled Artificial Intelligence (AI) to facilitate early stage drug development.

Early-stage drug discovery is highly dependent upon drug target evaluation, understanding of disease progression and identification of patient characteristics linked to disease progression overlaid upon chemical libraries of potential drug candidates. Artificial intelligence (AI) has become a credible approach towards dealing with the diversity and volume of data in the modern drug development phase. There are a growing number of services and solutions available to pharmaceutical sponsors though most prefer to constrain their own data to closed solutions given the intellectual property considerations. Newer platforms offer an alternative, outsourced solution leveraging sponsors data with other, external open-source data to anchor predictions (often proprietary algorithms) which are refined given data indexed upon the sponsor's own chemical libraries. Digital research environments (DREs) provide a mechanism to ingest, curate, integrate and otherwise manage the diverse data types relevant for drug discovery activities and also provide workspace services from which target sharing and collaboration can occur providing yet another alternative with sponsors being in control of the platform, data and predictive algorithms. Regulatory engagement will be essential in the operationalizing of the various solutions and alternatives; current treatment of drug discovery data may not be adequate with respect to both quality and useability in the future. More sophisticated AI/ML algorithms are likely based on current performance metrics and diverse data types (e.g., imaging and genomic data) will certainly be a more consistent part of the myriad of data types that fuel future AI-based algorithms. This favors a dynamic DRE-enabled environment to support drug discovery.

Target-driven machine learning-enabled virtual screening (TAME-VS) platform for early-stage hit identification.

High-throughput screening (HTS) methods enable the empirical evaluation of a large scale of compounds and can be augmented by virtual screening (VS) techniques to save time and money by using potential active compounds for experimental testing. Structure-based and ligand-based virtual screening approaches have been extensively studied and applied in drug discovery practice with proven outcomes in advancing candidate molecules. However, the experimental data required for VS are expensive, and hit identification in an effective and efficient manner is particularly challenging during early-stage drug discovery for novel protein targets. Herein, we present our TArget-driven Machine learning-Enabled VS (TAME-VS) platform, which leverages existing chemical databases of bioactive molecules to modularly facilitate hit finding. Our methodology enables bespoke hit identification campaigns through a user-defined protein target. The input target ID is used to perform a homology-based target expansion, followed by compound retrieval from a large compilation of molecules with experimentally validated activity. Compounds are subsequently vectorized and adopted for machine learning (ML) model training. These machine learning models are deployed to perform model-based inferential virtual screening, and compounds are nominated based on predicted activity. Our platform was retrospectively validated across ten diverse protein targets and demonstrated clear predictive power. The implemented methodology provides a flexible and efficient approach that is accessible to a wide range of users. The TAME-VS platform is publicly available at https://github.com/bymgood/Target-driven-ML-enabled-VS to facilitate early-stage hit identification.

PSnpBind-ML: predicting the effect of binding site mutations on protein-ligand binding affinity

Protein mutations, especially those which occur in the binding site, play an important role in inter-individual drug response and may alter binding affinity and thus impact the drug’s efficacy and side effects. Unfortunately, large-scale experimental screening of ligand-binding against protein variants is still time-consuming and expensive. Alternatively, in silico approaches can play a role in guiding those experiments. Methods ranging from computationally cheaper machine learning (ML) to the more expensive molecular dynamics have been applied to accurately predict the mutation effects. However, these effects have been mostly studied on limited and small datasets, while ideally a large dataset of binding affinity changes due to binding site mutations is needed. In this work, we used the PSnpBind database with six hundred thousand docking experiments to train a machine learning model predicting protein-ligand binding affinity for both wild-type proteins and their variants with a single-point mutation in the binding site. A numerical representation of the protein, binding site, mutation, and ligand information was encoded using 256 features, half of them were manually selected based on domain knowledge. A machine learning approach composed of two regression models is proposed, the first predicting wild-type protein-ligand binding affinity while the second predicting the mutated protein-ligand binding affinity. The best performing models reported an RMSE value within 0.5 - 0.6 kcal/mol-1 on an independent test set with an R2 value of 0.87 - 0.90. We report an improvement in the prediction performance compared to several reported models developed for protein-ligand binding affinity prediction. The obtained models can be used as a complementary method in early-stage drug discovery. They can be applied to rapidly obtain a better overview of the ligand binding affinity changes across protein variants carried by people in the population and narrow down the search space where more time-demanding methods can be used to identify potential leads that achieve a better affinity for all protein variants.

Insights into the Transport Cycle of LAT1 and Interaction with the Inhibitor JPH203.

The large Amino Acid Transporter 1 (LAT1) is an interesting target in drug discovery since this transporter is overexpressed in several human cancers. Furthermore, due to its location in the blood-brain barrier (BBB), LAT1 is interesting for delivering pro-drugs to the brain. In this work, we focused on defining the transport cycle of LAT1 using an in silico approach. So far, studies of the interaction of LAT1 with substrates and inhibitors have not considered that the transporter must undergo at least four different conformations to complete the transport cycle. We built outward-open and inward-occluded conformations of LAT1 using an optimized homology modelling procedure. We used these 3D models and the cryo-EM structures in outward-occluded and inward-open conformations to define the substrate/protein interaction during the transport cycle. We found that the binding scores for the substrate depend on the conformation, with the occluded states as the crucial steps affecting the substrate affinity. Finally, we analyzed the interaction of JPH203, a high-affinity inhibitor of LAT1. The results indicate that conformational states must be considered for in silico analyses and early-stage drug discovery. The two built models, together with the available cryo-EM 3D structures, provide important information on the LAT1 transport cycle, which could be used to speed up the identification of potential inhibitors through in silico screening.

Industry update, December 2022.

Industry update, December 2022.

Pose classification using three-dimensional atomic structure-based neural networks applied to ion channel-ligand docking.

The identification of promising lead compounds showing pharmacological activities toward a biological target is essential in early stage drug discovery. With the recent increase in available small-molecule databases, virtual high-throughput screening using physics-based molecular docking has emerged as an essential tool in assisting fast and cost-efficient lead discovery and optimization. However, the best scored docking poses are often suboptimal, resulting in incorrect screening and chemical property calculation. We address the pose classification problem by leveraging data-driven machine learning approaches to identify correct docking poses from AutoDock Vina and Glide screens. To enable effective classification of docking poses, we present two convolutional neural network approaches: a three-dimensional convolutional neural network (3D-CNN) and an attention-based point cloud network (PCN) trained on the PDBbind refined set. We demonstrate the effectiveness of our proposed classifiers on multiple evaluation data sets including the standard PDBbind CASF-2016 benchmark data set and various compound libraries with structurally different protein targets including an ion channel data set extracted from Protein Data Bank (PDB) and an in-house KCa3.1 inhibitor data set. Our experiments show that excluding false positive docking poses using the proposed classifiers improves virtual high-throughput screening to identify novel molecules against each target protein compared to the initial screen based on the docking scores.

Development of machine learning models for prediction of antibody non-specificity.

For the development of therapeutic antibodies, it is acknowledged that optimal target binding affinity should be accompanied by high specificity. In recent years, there has been an increased focus on lowering non-specific binding during early-stage drug development. In particular, non-specificity has been recognised as a root-cause for failure in numerous antibody-based drug programs due to unexpected pharmacokinetics and elevated toxicity. Hence, there is a need for predictive tools for flagging and deselecting non-specific antibodies at an early stage to mitigate failure at clinical late-stage development. From a computational design perspective, prediction has remained a challenge as non-specificity is currently a poorly understood biophysical phenomena, typically involving weak interactions to various, unrelated targets and surfaces. In this work, we describe the development of sequence-based machine learning (ML) models for prediction of antibody non-specificity. A dataset of >1000 antibody sequences paired with their respective non-specificity flag, defined by ELISA using a panel of common antigens, were retrieved from public sources (Jain et al., 2017; Boughter et al., 2020). The sequences were embedded using state-of-the-art protein language models for representation of their physiochemical properties, whereafter the vectorized embeddings were served as features for training of binary classifiers for non-specificity flags. A total of 12 different antibody domain-specific models were generated. Highest predictability of non-specificity was obtained for ML algorithms trained on the heavy-chain variable domain (VH), and the complementarity determining regions (CDRs), and resulted in a mean accuracy of 73% across 3-, 5-, and 10-fold cross-validations. The classification probability of non-specificity followed an uptrend with the experimental data (non-specificity flags), thereby mimicking regression behaviour. Altogether, our work demonstrates feasibility for prediction of non-specific binding from the antibody sequence, and that the non-specificity primarily stems from the VH domain, with main contributions from the heavy-chain CDRs.

Electrochemical biosensors based on saliva electrolytes for rapid detection and diagnosis.

In recent years, electrochemical biosensors (ECBSs) have shown significant potential for real-time disease diagnosis and in situ physical condition monitoring. As a multi-constituent oral fluid comprising various disease signaling biomarkers, saliva has drawn much attention in the field of point-of-care (POC) testing. In particular, during the outbreak of the COVID-19 pandemic, ECBSs which hold the simplicity of a single-step assay compared with the multi-step assay of traditional testing methods are expected to relieve the human and economic burden caused by the massive and long-term sample testing process. Noteworthily, ECBSs for the detection of SARS-CoV-2 in saliva have already been developed and may replace current testing methods. Furthermore, the detection scope has expanded from routine indices such as sugar and uric acid to abnormal biomarkers for early-stage disease detection and drug level monitoring, which further facilitated the evolution of ECBSs in the last 5 years. This review is divided into several main sections. First, we discussed the latest advancements and representative research on ECBSs for saliva testing. Then, we focused on a novel kind of ECBS, organic electrochemical transistors (OECTs), which hold great advantages of high sensitivity and signal-to-noise ratio and on-site detection. Finally, application of ECBSs with integrated portable platforms in oral cavities, which lead to powerful auxiliary testing means for telemedicine, has also been discussed.

Exploring Anticancer Potential in Bajakah Tampala by In Silico Virtual Screening

Incidences of cancer have increased sharply throughout the world. The process of finding a targeted cancer drug takes a long time and a lot of money. One method that can help overcome this is computational methods, such as virtual screening. It can be an important alternative in early-stage drug discovery. The aim of this study is to explore and analyze the potential protein targets of Bajakah tampala (Spatholobulus littoralis) as an anticancer chemotherapy. This is a bioinformatics study that uses the in silico method through the pathway analysis method with PubChem software, Swiss Target prediction, String and Cytoscope. The results showed that many phytochemicals present in Spatholobulus littoralis Hask are predicted to be anticancer.
 Keywords: Anticancer, Bajakah tampala, In silico, Spatholobulus littoralis

Advances in luminescence-based technologies for drug discovery

ABSTRACT Introduction Luminescence-based technologies, specifically bioluminescence and chemiluminescence, are powerful tools with extensive use in drug discovery. Production of light during chemiluminescence and bioluminescence, unlike fluorescence, doesn’t require an excitation light source, resulting in high signal-to-noise ratio, less background interference, and no issues from phototoxicity and photobleaching. These characteristics of luminescence technologies offer unique advantages for experimental designs, allowing for greater flexibility to target a wide range of proteins and biological processes for drug discovery at different stages. Areas covered This review provides a basic overview of luciferase-based technologies and details recent advances and use cases of luciferase and luciferin variations and their applicability in the drug discovery toolset. The authors expand upon specific applications of luciferase technologies, including chemiluminescent and bioluminescent-based microscopy. Finally, the authors lay out forward-looking statements on the field of luminescence and how it may shape the translational scientists’ work moving forward. Expert opinion The demand for improved luciferase and luciferin pairs correlates strongly with efforts to improve the sensitivity and robustness of high-throughput assays. As luminescent reporter systems improve, so will the expansion of use cases for luminescence-based technologies in early-stage drug discovery. With the synthesis of novel, non-enzymatic chemiluminescence-based probes, which previously were restrained to only basic research applications, they may now be readily implemented in drug discovery campaigns.

Effect of Binding Linkers on the Efficiency and Metabolite Profile of Biomimetic Reactions Catalyzed by Immobilized Metalloporphyrin.

The investigation of liver-related metabolic stability of a drug candidate is a widely used key strategy in early-stage drug discovery. Metalloporphyrin-based biomimetic catalysts are good and well-described models of the function of CyP450 in hepatocytes. In this research, the immobilization of an iron porphyrin was performed on nanoporous silica particles via ionic interactions. The effect of the metalloporphyrin binding linkers was investigated on the catalytic efficiency and the metabolic profile of chloroquine as a model drug. The length of the amino-substituted linkers affects the chloroquine conversion as well as the ratio of human major and minor metabolites. While testing the immobilized catalysts in the continuous-flow reactor, results showed that the presented biomimetic system could be a promising alternative for the early-stage investigation of drug metabolites regarding analytical or synthetic goals as well.

Translating atherosclerosis research from bench to bedside: navigating the barriers for effective preclinical drug discovery.

Cardiovascular disease (CVD) remains the leading cause of death worldwide. An ongoing challenge remains the development of novel pharmacotherapies to treat CVD, particularly atherosclerosis. Effective mechanism-informed development and translation of new drugs requires a deep understanding of the known and currently unknown biological mechanisms underpinning atherosclerosis, accompanied by optimization of traditional drug discovery approaches. Current animal models do not precisely recapitulate the pathobiology underpinning human CVD. Accordingly, a fundamental limitation in early-stage drug discovery has been the lack of consensus regarding an appropriate experimental in vivo model that can mimic human atherosclerosis. However, when coupled with a clear understanding of the specific advantages and limitations of the model employed, preclinical animal models remain a crucial component for evaluating pharmacological interventions. Within this perspective, we will provide an overview of the mechanisms and modalities of atherosclerotic drugs, including those in the preclinical and early clinical development stage. Additionally, we highlight recent preclinical models that have improved our understanding of atherosclerosis and associated clinical consequences and propose model adaptations to facilitate the development of new and effective treatments.

Assessing How Residual Errors of Scoring Functions Correlate to Ligand Structural Features

Scoring functions (SFs) are ubiquitous tools for early stage drug discovery. However, their accuracy currently remains quite moderate. Despite a number of successful target-specific SFs appearing recently, up until now, no ideas on how to systematically improve the general scope of SFs have been formulated. In this work, we hypothesized that the specific features of ligands, corresponding to interactions well appreciated by medicinal chemists (e.g., hydrogen bonds, hydrophobic and aromatic interactions), might be responsible, in part, for the remaining SF errors. The latter provides direction to efforts aimed at the rational and systematic improvement of SF accuracy. In this proof-of-concept work, we took a CASF-2016 coreset of 285 ligands as a basis for comparison and calculated the values of scores for a representative panel of SFs (including AutoDock 4.2, AutoDock Vina, X-Score, NNScore2.0, ΔVina RF20, and DSX). The residual error of linear correlation of each SF value, with the experimental values of affinity and activity, was then analyzed in terms of its correlation with the presence of the fragments responsible for certain medicinal chemistry defined interactions. We showed that, despite the fact that SFs generally perform reasonably, there is room for improvement in terms of better parameterization of interactions involving certain fragments in ligands. Thus, this approach opens a potential way for the systematic improvement of SFs without their significant complication. However, the straightforward application of the proposed approach is limited by the scarcity of reliable available data for ligand-receptor complexes, which is a common problem in the field.

Novel biomimetic nanocomposite for investigation of drug metabolism

In vitro mimicking of hepatic drug metabolism is a key issue in early-stage drug discovery. Synthetic metalloporphyrins show structural similarity with the heme type prosthetic group of cytochrome P450 as primary hepatic enzyme in oxidative drug biotransformation. Therefore, they can catalyze these oxidations. Concerning economical aspects and the poor stability of metalloporphyrin, their immobilization onto or into solid carriers can be promising solution. This study presents a novel immobilized metalloporphyrin nanocomposite system and its potential use as biomimetic catalysts. The developed two-step immobilization procedure consists of two main steps. First, the ionic binding of meso-tetra(parasulphonatophenyl) iron porphyrin onto functionalized magnetic nanoparticles is established, followed by embedding the nanoparticles into polylactic acid nanofibers by electrospinning technique. Due to the synergistic morphological and chemo-structural advantages of binding onto nanoparticles and embedding in polymeric matrices the biomimetic efficiency of metalloporphyrin can be remarkably enhanced, while substrate conversion value was tenfold larger than which could be achieved with classic human liver microsomal system.

Characterization of KLH-driven immune responses in clinical studies: A systematic review

The pharmacological activity assessment of novel immunomodulatory drugs in early-stage drug development is challenging as healthy volunteers do not express relevant immune biomarkers. Alternatively, the immune system can be challenged with keyhole limpet hemocyanin (KLH), a suitable antigen for studying adaptive immune responses. This report systemically reviews the KLH challenge in clinical studies focusing on the characterization of the KLH-driven systemic and local immune responses, identification of the KLH-induced biomarkers, and the evaluation of the effect of pharmacological interventions and diseases on the KLH response. A systematic literature review was carried out in PubMed spanning from 1967 to 2022. The systemic humoral KLH responses could be characterized by ELISA after 3 weeks following immunization. For the systemic cellular and molecular immune responses multiple KLH immunizations and the use of novel techniques such as flow cytometry and ELISpot yield optimal results. The objective evaluation of dermal KLH rechallenge allows for more accurate and sensitive quantification of the local response compared to subjective scoring. For the local cellular and molecular assays after KLH dermal rechallenge we also advocate the use of multiple KLH immunizations. Furthermore, oral KLH feeding, age, physical activity, alcohol consumption, stress, as well as certain auto-immune diseases also play a role in the KLH-induced immune response. Importantly, based on the KLH challenges, the effect of (novel) immunomodulatory drugs could be demonstrated in healthy volunteers, providing valuable information for the clinical development of these compounds. This review underlines the value of KLH challenges in clinical studies, but also the need for standardized and well-controlled methodology to induce and evaluate KLH responses.Systematic Review Registration:https://www.crd.york.ac.uk/prospero/, identifierCRD42022335419

Large-Scale Crystallographic Fragment Screening Expedites Compound Optimization and Identifies Putative Protein-Protein Interaction Sites.

The identification of starting points for compound development is one of the key steps in early-stage drug discovery. Information-rich techniques such as crystallographic fragment screening can potentially increase the efficiency of this step by providing the structural information of the binding mode of the ligands in addition to the mere binding information. Here, we present the crystallographic screening of our 1000-plus-compound F2X-Universal Library against the complex of the yeast spliceosomal Prp8 RNaseH-like domain and the snRNP assembly factor Aar2. The observed 269 hits are distributed over 10 distinct binding sites on the surface of the protein-protein complex. Our work shows that hit clusters from large-scale crystallographic fragment screening campaigns identify known interaction sites with other proteins and suggest putative additional interaction sites. Furthermore, the inherent binding pose validation within the hit clusters may accelerate downstream compound optimization.

The Real Effects of Equity Markets on Innovation

AbstractIn theory, financial markets promote innovation by selectively allocating capital to high-quality projects. In this article, I show that equity markets can also inhibit innovation. In public firms, I find that short-term equity market declines cause pharmaceutical companies to abandon early-stage drug developments, irrespective of drug quality or changes in a firm’s stock price. I show that financing constraints drive this behavior, highlighting that even short-term market fluctuations can have long-term effects on pharmaceutical innovation and prevent potentially life-saving drugs from progressing to the market.

Random-forest model for drug–target interaction prediction via Kullback–Leibler divergence

Virtual screening has significantly improved the success rate of early stage drug discovery. Recent virtual screening methods have improved owing to advances in machine learning and chemical information. Among these advances, the creative extraction of drug features is important for predicting drug–target interaction (DTI), which is a large-scale virtual screening of known drugs. Herein, we report Kullback–Leibler divergence (KLD) as a DTI feature and the feature-driven classification model applicable to DTI prediction. For the purpose, E3FP three-dimensional (3D) molecular fingerprints of drugs as a molecular representation allow the computation of 3D similarities between ligands within each target (Q–Q matrix) to identify the uniqueness of pharmacological targets and those between a query and a ligand (Q–L vector) in DTIs. The 3D similarity matrices are transformed into probability density functions via kernel density estimation as a nonparametric estimation. Each density model can exploit the characteristics of each pharmacological target and measure the quasi-distance between the ligands. Furthermore, we developed a random forest model from the KLD feature vectors to successfully predict DTIs for representative 17 targets (mean accuracy: 0.882, out-of-bag score estimate: 0.876, ROC AUC: 0.990). The method is applicable for 2D chemical similarity.

DENVIS: Scalable and High-Throughput Virtual Screening Using Graph Neural Networks with Atomic and Surface Protein Pocket Features.

Computational methods for virtual screening can dramatically accelerate early-stage drug discovery by identifying potential hits for a specified target. Docking algorithms traditionally use physics-based simulations to address this challenge by estimating the binding orientation of a query protein-ligand pair and a corresponding binding affinity score. Over the recent years, classical and modern machine learning architectures have shown potential for outperforming traditional docking algorithms. Nevertheless, most learning-based algorithms still rely on the availability of the protein-ligand complex binding pose, typically estimated via docking simulations, which leads to a severe slowdown of the overall virtual screening process. A family of algorithms processing target information at the amino acid sequence level avoid this requirement, however, at the cost of processing protein data at a higher representation level. We introduce deep neural virtual screening (DENVIS), an end-to-end pipeline for virtual screening using graph neural networks (GNNs). By performing experiments on two benchmark databases, we show that our method performs competitively to several docking-based, machine learning-based, and hybrid docking/machine learning-based algorithms. By avoiding the intermediate docking step, DENVIS exhibits several orders of magnitude faster screening times (i.e., higher throughput) than both docking-based and hybrid models. When compared to an amino acid sequence-based machine learning model with comparable screening times, DENVIS achieves dramatically better performance. Some key elements of our approach include protein pocket modeling using a combination of atomic and surface features, the use of model ensembles, and data augmentation via artificial negative sampling during model training. In summary, DENVIS achieves competitive to state-of-the-art virtual screening performance, while offering the potential to scale to billions of molecules using minimal computational resources.