Drug Discovery Scientists Research Articles

LagNet: Deep Lagrangian Mechanics for Plug-and-Play Molecular Representation Learning

Molecular representation learning is a fundamental problem in the field of drug discovery and molecular science. Whereas incorporating molecular 3D information in the representations of molecule seems beneficial, which is related to computational chemistry with the basic task of predicting stable 3D structures (conformations) of molecules. Existing machine learning methods either rely on 1D and 2D molecular properties or simulate molecular force field to use additional 3D structure information via Hamiltonian network. The former has the disadvantage of ignoring important 3D structure features, while the latter has the disadvantage that existing Hamiltonian neural network must satisfy the “canonial” constraint, which is difficult to be obeyed in many cases. In this paper, we propose a novel plug-and-play architecture LagNet by simulating molecular force field only with parameterized position coordinates, which implements Lagrangian mechanics to learn molecular representation by preserving 3D conformation without obeying any additional restrictions. LagNet is designed to generate known conformations and generalize for unknown ones from molecular SMILES. Implicit positions in LagNet are learned iteratively using discrete-time Lagrangian equations. Experimental results show that LagNet can well learn 3D molecular structure features, and outperforms previous state-of-the-art baselines related molecular representation by a significant margin.

Conditional Diffusion Based on Discrete Graph Structures for Molecular Graph Generation

Learning the underlying distribution of molecular graphs and generating high-fidelity samples is a fundamental research problem in drug discovery and material science. However, accurately modeling distribution and rapidly generating novel molecular graphs remain crucial and challenging goals. To accomplish these goals, we propose a novel Conditional Diffusion model based on discrete Graph Structures (CDGS) for molecular graph generation. Specifically, we construct a forward graph diffusion process on both graph structures and inherent features through stochastic differential equations (SDE) and derive discrete graph structures as the condition for reverse generative processes. We present a specialized hybrid graph noise prediction model that extracts the global context and the local node-edge dependency from intermediate graph states. We further utilize ordinary differential equation (ODE) solvers for efficient graph sampling, based on the semi-linear structure of the probability flow ODE. We also combine the solvers with gradient guidance from the molecule property predictor for similarity-constrained molecule optimization. Experiments on diverse datasets validate the effectiveness of our framework. Particularly, the proposed method still generates high-quality molecular graphs in a limited number of steps.

Faster and more diverse de novo molecular optimization with double-loop reinforcement learning using augmented SMILES.

Using generative deep learning models and reinforcement learning together can effectively generate new molecules with desired properties. By employing a multi-objective scoring function, thousands of high-scoring molecules can be generated, making this approach useful for drug discovery and material science. However, the application of these methods can be hindered by computationally expensive or time-consuming scoring procedures, particularly when a large number of function calls are required as feedback in the reinforcement learning optimization. Here, we propose the use of double-loop reinforcement learning with simplified molecular line entry system (SMILES) augmentation to improve the efficiency and speed of the optimization. By adding an inner loop that augments the generated SMILES strings to non-canonical SMILES for use in additional reinforcement learning rounds, we can both reuse the scoring calculations on the molecular level, thereby speeding up the learning process, as well as offer additional protection against mode collapse. We find that employing between 5 and 10 augmentation repetitions is optimal for the scoring functions tested and is further associated with an increased diversity in the generated compounds, improved reproducibility of the sampling runs and the generation of molecules of higher similarity to known ligands.

Synthesis of 2-Formyl Carbazoles via Tandem Reaction of Indolyl Nitrones with 2-Methylidene Cyclic Carbonate.

The synthesis of functionalized carbazoles as privileged nitrogen heterocycles has emerged as a central topic in drug discovery and material science. We herein disclose the rhodium(III)-catalyzed cross-coupling reaction between indolyl nitrones and 2-methylidene cyclic carbonate as an allylating surrogate, resulting in the formation of C2-formylated carbazoles via tandem C-H allylation, [3 + 2] cycloaddition, aromatization, and benzylic oxidation. The synthetic utility of this protocol is highlighted by a variety of post-transformations of C2-formylated carbazoles.

SELFormer: molecular representation learning via SELFIES language models

Automated computational analysis of the vast chemical space is critical for numerous fields of research such as drug discovery and material science. Representation learning techniques have recently been employed with the primary objective of generating compact and informative numerical expressions of complex data, for efficient usage in subsequent prediction tasks. One approach to efficiently learn molecular representations is processing string-based notations of chemicals via natural language processing algorithms. Majority of the methods proposed so far utilize SMILES notations for this purpose, which is the most extensively used string-based encoding for molecules. However, SMILES is associated with numerous problems related to validity and robustness, which may prevent the model from effectively uncovering the knowledge hidden in the data. In this study, we propose SELFormer, a transformer architecture-based chemical language model (CLM) that utilizes a 100% valid, compact and expressive notation, SELFIES, as input, in order to learn flexible and high-quality molecular representations. SELFormer is pre-trained on two million drug-like compounds and fine-tuned for diverse molecular property prediction tasks. Our performance evaluation has revealed that, SELFormer outperforms all competing methods, including graph learning-based approaches and SMILES-based CLMs, on predicting aqueous solubility of molecules and adverse drug reactions, while producing comparable results for the remaining tasks. We also visualized molecular representations learned by SELFormer via dimensionality reduction, which indicated that even the pre-trained model can discriminate molecules with differing structural properties. We shared SELFormer as a programmatic tool, together with its datasets and pre-trained models at https://github.com/HUBioDataLab/SELFormer. Overall, our research demonstrates the benefit of using the SELFIES notations in the context of chemical language modeling and opens up new possibilities for the design and discovery of novel drug candidates with desired features.

Unveiling the Stereoselectivity and Regioselectivity of the [3+2] Cycloaddition Reaction between N-methyl-C-4-methylphenyl-nitrone and 2-Propynamide from a MEDT Perspective.

[3+2] cycloaddition reactions play a crucial role in synthesizing complex organic molecules and have significant applications in drug discovery and materials science. In this study, the [3+2] cycloaddition (32CA) reactions of N-methyl-C-4-methyl phenyl-nitrone 1 and 2-propynamide 2, which have not been extensively studied before, were investigated using molecular electron density theory (MEDT) at the B3LYP/6-311++G(d,p) level of theory. According to an electron localization function (ELF) study, N-methyl-C-4-methyl phenyl-nitrone 1 is a zwitterionic species with no pseudoradical or carbenoid centers. Conceptual density functional theory (CDFT) indices were used to predict the global electronic flux from the strong nucleophilic N-methyl-C-4-methyl phenylnitrone 1 to the electrophilic 2-propynamide 2 functions. The 32CA reactions proceeded through two pairs of stereo- and regioisomeric reaction pathways to generate four different products: 3, 4, 5, and 6. The reaction pathways were irreversible owing to their exothermic characters: -136.48, -130.08, -130.99, and -140.81 kJ mol-1, respectively. The enthalpy of the 32CA reaction leading to the formation of cycloadduct 6 was lower compared with the other path owing to a slight increase in its polar character, observed through the global electron density transfer (GEDT) during the transition states and along the reaction path. A bonding evolution theory (BET) analysis showed that these 32CA reactions proceed through the coupling of pseudoradical centers, and the formation of new C-C and C-O covalent bonds did not begin in the transition states.

Chirality Sensing of N-Heterocycles via 19F NMR.

Methods to rapidly detect and differentiate chiral N-heterocyclic compounds become increasingly important owing to the widespread application of N-heterocycles in drug discovery and materials science. We herein report a 19F NMR-based chemosensing approach for the prompt enantioanalysis of various N-heterocycles, where the dynamic binding between the analytes and a chiral 19F-labeled palladium probe create characteristic 19F NMR signals assignable to each enantiomer. The open binding site of the probe allows the effective recognition of bulky analytes that are otherwise difficult to detect. The chirality center distal to the binding site is found sufficient for the probe to discriminate the stereoconfiguration of the analyte. The utility of the method in the screening of reaction conditions for the asymmetric synthesis of lansoprazole is demonstrated.

Advanced modeling based on machine learning for evaluation of drug nanoparticle preparation via green technology: Theoretical assessment of solubility variations

Variability of drug solubility in supercritical solvent was investigated theoretically in this study using multiple machine learning models. The method of supercritical preparation can be used to manufacture solid-dosage drugs with enhanced solubility which can increase drug bioavailability and efficacy. In this study, we applied bagging regression to three core models, namely SVM, KNN, and linear regression (LR) to predict the solubility and density of Hyoscine drug based on Temperature and Pressure parameters as inputs. The WOA optimization algorithm was used to tune hyperparameters, and the performance of these models was compared with r2, RMSE, MAE, and AARD% metrics. The results revealed that the BAG-KNN model outperformed the others in predicting solubility, with an r2 score of 0.989 and a low RMSE of 0.041. Also, the BAG-LR model had the highest AARD% value of 7.50974, indicating that it has higher average relative deviation from the actual density values. Overall, the bagging approach proved to be effective in improving the accuracy of the core models, and the WOA optimization algorithm was successful in fine-tuning their hyperparameters. These findings can have significant implications for the accurate prediction of chemical properties in various fields, such as drug discovery and material science.

N‐(Morpholine‐4‐dithio)phthalimide: A Shelf‐Stable, Bilateral Platform Molecule Enabling Access to Diverse Unsymmetrical Disulfides**

AbstractSynthetic methods for unsymmetrical disulfides are greatly needed owing to their applications in drug discovery, linker chemistry, and materials sciences. In this study, a new shelf‐stable and easy‐to‐prepare bilateral disulfurating platform molecule, N‐(morpholine‐4‐dithio)phthalimide, has been developed for the divergent synthesis of unsymmetrical disulfides. The amino and imide leaving groups of this reagent can be orthogonally transformed. Under acidic conditions, the amino moiety undergoes selective protonation and thus can be displaced by various carbon nucleophiles, such as allyl trimethylsilanes, alkynyl silanes, and electron‐rich arenes, leaving the phthalimide moiety untouched. Meanwhile, the phthalimide leaving group is amenable to substitution under basic or neutral conditions. The combination of these transformations provides rapid access to diverse unsymmetrical disulfides through two C−S bond‐forming reactions.

N-(Morpholine-4-dithio)phthalimide: A Shelf-Stable, Bilateral Platform Molecule Enabling Access to Diverse Unsymmetrical Disulfides.

Synthetic methods for unsymmetrical disulfides are greatly needed owing to their applications in drug discovery, linker chemistry, and materials sciences. In this study, a new shelf-stable and easy-to-prepare bilateral disulfurating platform molecule, N-(morpholine-4-dithio)phthalimide, has been developed for the divergent synthesis of unsymmetrical disulfides. The amino and imide leaving groups of this reagent can be orthogonally transformed. Under acidic conditions, the amino moiety undergoes selective protonation and thus can be displaced by various carbon nucleophiles, such as allyl trimethylsilanes, alkynyl silanes, and electron-rich arenes, leaving the phthalimide moiety untouched. Meanwhile, the phthalimide leaving group is amenable to substitution under basic or neutral conditions. The combination of these transformations provides rapid access to diverse unsymmetrical disulfides through two C-S bond-forming reactions.

Stepwise on-demand functionalization of multihydrosilanes enabled by a hydrogen-atom-transfer photocatalyst based on eosin Y

Organosilanes are of vital importance for modern human society, having found widespread applications in functional materials, organic synthesis, drug discovery and life sciences. However, their preparation remains far from trivial, and on-demand synthesis of heteroleptic substituted silicon reagents is a formidable challenge. The generation of silyl radicals from hydrosilanes via direct hydrogen-atom-transfer (HAT) photocatalysis represents the most atom-, step-, redox- and catalyst-economic pathway for the activation of hydrosilanes. Here, in view of the green characteristics of neutral eosin Y (such as its abundance, low cost, metal-free nature, absorption of visible light and excellent selectivity), we show that using it as a direct HAT photocatalyst enables the stepwise custom functionalization of multihydrosilanes, giving access to fully substituted silicon compounds. By exploiting this strategy, we realize preferable hydrogen abstraction of Si-H bonds in the presence of active C-H bonds, diverse functionalization of hydrosilanes (for example, alkylation, vinylation, allylation, arylation, deuteration, oxidation and halogenation), and remarkably selective monofunctionalization of di- and trihydrosilanes.

Microfluidic Chip with Fiber-Tip Sensors for Synchronously Monitoring Concentration and Temperature of Glucose Solutions

Monitoring the properties of fluids in microfluidic chips often requires complex open-space optics technology and expensive equipment. In this work, we introduce dual-parameter optical sensors with fiber tips into the microfluidic chip. Multiple sensors were distributed in each channel of the chip, which enabled the real-time monitoring of the concentration and temperature of the microfluidics. The temperature sensitivity and glucose concentration sensitivity could reach 314 pm/°C and -0.678 dB/(g/L), respectively. The hemispherical probe hardly affected the microfluidic flow field. The integrated technology combined the optical fiber sensor with the microfluidic chip and was low cost with high performance. Therefore, we believe that the proposed microfluidic chip integrated with the optical sensor is beneficial for drug discovery, pathological research and material science investigation. The integrated technology has great application potential for micro total analysis systems (μ-TAS).

Structural analysis and ensemble docking revealed the binding modes of selected progesterone receptor modulators

Uterine fibroids (UF) are benign smooth muscle neoplasm of uterus that have a significant impact on a woman’s quality of life as they perturb hormonal homeostasis resulting in heavy menstrual bleeding, impaired fertility, pregnancy complications and loss. UF can be surgically removed through invasive procedures, but their recurrence rate is often high. Progesterone receptor (PR) has an imperative role in UF management. Mifepristone, ulipristal acetate (UPA) and asoprisnil (ASO) are some selective progesterone receptor modulators (SPRMs), acts on PR, but due to their side effects in long term use, they were withdrawn from the market. Hence, there is a dire need for novel, highly efficient with least side effects, therapeutics for the treatment of UF. To contribute toward the drug discovery for UF, we made an extensive structural comparison of reported PR crystal structures, also elucidated the binding modes of four existing SPRMs against human PR through ensemble docking approach. Our studies revealed the presence of 5 highly repeating water molecules that has an important role in ligand binding and structural stability. Our ensemble docking and MD simulation revealed that studied ligands have preferential selectivity toward the specific conformation of PR. It is anticipated that our study will be a useful resource to all the drug discovery scientists who are engaged in the identification of lead molecules against UF.

Machine learning models to accelerate the design of polymeric long-acting injectables

Long-acting injectables are considered one of the most promising therapeutic strategies for the treatment of chronic diseases as they can afford improved therapeutic efficacy, safety, and patient compliance. The use of polymer materials in such a drug formulation strategy can offer unparalleled diversity owing to the ability to synthesize materials with a wide range of properties. However, the interplay between multiple parameters, including the physicochemical properties of the drug and polymer, make it very difficult to intuitively predict the performance of these systems. This necessitates the development and characterization of a wide array of formulation candidates through extensive and time-consuming in vitro experimentation. Machine learning is enabling leap-step advances in a number of fields including drug discovery and materials science. The current study takes a critical step towards data-driven drug formulation development with an emphasis on long-acting injectables. Here we show that machine learning algorithms can be used to predict experimental drug release from these advanced drug delivery systems. We also demonstrate that these trained models can be used to guide the design of new long acting injectables. The implementation of the described data-driven approach has the potential to reduce the time and cost associated with drug formulation development.

Characteristic features and comparative analysis of essential oil composition of selected genus of Ocimum sanctum L. through GC-MS.

The main aim of this research is to provide literature on the Ocimum plant, and to know the significance of the Ocimum species carried out by pharmacognostic study and experimental design for GC-MS. Ocimum genus are very important for their therapeutic potential among the most important aromatic herbs. Extreme attention has been put on literature reports in which the utilization of tulsi and their pharmacognostic study has been done by performing morphological and microscopic leaf experimental design and by using essential oil through the GC-MS instrumentation method. The utilization of these characteristics would be important for the drug discovery scientist to develop a specific formulation of the crude drug, which will be a magical therapeutic agent in the future, with many advantages. GC-MS chromatogram of Ocimum sanctum, Ocimum canum, and Ocimum gratissimum oil showed major peaks and has been identified after comparison of the mass spectra with the NIST library, indicating the presence of three phytocomponents. From the results, the GC-MS study suggested that anethole which is well reported antimicrobial compound is more in O. canum (2.66%) in comparison to O. sanctum (1,28%) but absent in O. gratissimum. The results indicated that the antimicrobial activity is more in O. canum due to the presence ofa high amount of anethole in comparison to O. gratissimum and O. sanctum. The result revealed that O. canum has a microscopic character that can be identified by the characteristic GC MS analysis of extracts to distinguish between different species of the ocimum plant.

Challenges in natural product-based drug discovery assisted with in silico-based methods.

The application of traditional medicine by humans for the treatment of ailments as well as improving the quality of life far outdates recorded history. To date, a significant percentage of humans, especially those living in developing/underprivileged communities still rely on traditional medicine for primary healthcare needs. In silico-based methods have been shown to play a pivotal role in modern pharmaceutical drug discovery processes. The application of these methods in identifying natural product (NP)-based hits has been successful. This is very much observed in many research set-ups that use rationally in silico-based methods in combination with experimental validation techniques. The combination has rendered the use of in silico-based approaches even more popular and successful in the investigation of NPs. However, identifying and proposing novel NP-based hits for experimental validation comes with several challenges such as the availability of compounds by suppliers, the huge task of separating pure compounds from complex mixtures, the quantity of samples available from the natural source to be tested, not to mention the potential ecological impact if the natural source is exhausted. Because most peer-reviewed publications are biased towards "positive results", these challenges are generally not discussed in publications. In this review, we highlight and discuss these challenges. The idea is to give interested scientists in this field of research an idea of what they can come across or should be expecting as well as prompting them on how to avoid or fix these issues.

UV-mediated enhancement of antibacterial secondary metabolites in endophytic Lasiodiplodia theobromae

In the science of drug discovery, ultraviolet (UV) irradiation has been applied to induce mutagenesis in fungi to provide possibilities for the stimulation or enhancement of fungal biosynthetic capabilities. This study was carried out to evaluate the effect of UV radiation on the biosynthesis of antibacterial secondary metabolites in an endophytic Lasiodiplodia theobromae. Using standard methods, the fungus was isolated from healthy leaves of Cola acuminata and identified based on PCR amplification and genomic sequencing of the internal transcribed spacer (ITS) region. Cultures of L. theobromae were exposed to UV radiation at different time intervals of 1, 2 and 5 min. The fungus was subjected to solid-state fermentation in rice medium before and after UV treatments. The fungal secondary metabolites were extracted and tested for antibacterial activity using the agar diffusion method. Compounds present in the obtained extracts were identified by HPLC and GC-MS analysis. At a concentration of 1 mg/ml, the extract of the wild type L. theobromae (untreated) was observed to only inhibit Staphylococcus aureus, with an IZD of 12 mm. However, the extract of UV-treated L. theobromae (2 min) inhibited S. aureus, Escherichia coli and Pseudomonas aeruginosa with an IZD of 10 and 4 mm respectively. A wide array of compounds in the phenolics, fatty acids, alkaloids and alkanes classes were identified in the UV-treated and untreated fungal extracts. Overall, UV treatments of L. theobromae stimulated the production of seventeen (17) new compounds that were not detected in the untreated strain. The study confirms UV irradiation as an effective method for stimulating microbial biosynthesis of new bioactive compounds, indicating a promising and potentially abundant source of new drug compounds from microorganisms.

Improved Synthesis of the Anti-SARS-CoV-2 Investigational Agent (E)-N-(4-Cyanobenzylidene)-6-fluoro-3-hydroxypyrazine-2-carboxamide (Cyanorona-20)

Medicinal chemistry scientists` efforts and trials to discover a very potent anticoronaviral medicine specifically effective against the current frightening virus, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), are not over yet. Synthetic organic chemistry will remain one of the most important branches in the entire drug discovery science. (E)-N-(4-Cyanobenzylidene)-6-fluoro-3-hydroxypyrazine-2-carboxamide (Cyanorona-20), a newly-discovered favipiravir analog/ derivative, is one of the promising synthetic organic compounds that displayed very strong nanomolar potencies against this fatal coronavirus, reaching an anticoronaviral-2 EC50 of nearly 450 nM or 0.45 μM. This compound was found to act against the SARS-CoV-2 mainly through the powerful inhibition of the coronaviral RNA-dependent RNA polymerase (RdRp), via competitively occupying and locking this enzyme`s major catalytic active site pocket (the suggested primary mechanism of action). Cyanorona-20 is still under progressive investigation as an attempt to continue developing it as a prospective remedy for the coronavirus disease 2019 (COVID-19). However, the previous literature synthetic procedures of Cyanorona-20 were criticized for several reasons like the harsh handling, difficult separation, small yield, and low purity. Herein in this short-communication or technical-note article, more reproducible and efficient novel synthetic method for Cyanorona-20 compound is presented, in an effort to address almost all of the problems which were accompanying the preceding methods.

Temozolomide Sensitivity of Malignant Glioma Cell Lines – A Systematic Review Assessing Consistencies Between In Vitro Studies

Abstract AIMS Malignant glioma cell line models are integral to pre-clinical testing of novel potential therapies. Accurate prediction of likely efficacy in the clinic requires that these models are reliable and consistent. We assessed this by examining the reporting of experimental conditions and sensitivity to temozolomide in glioma cells lines. METHOD We searched Medline and Embase (Jan 1994-Jan 2021) for studies evaluating the effect of temozolomide monotherapy on cell viability of at least one malignant glioma cell line. Key data items included type of cell lines, temozolomide exposure duration in hours (hr), and cell viability measure (IC50). RESULTS We included 212 studies from 2789 non-duplicate records that reported 248 distinct cell lines. The commonest cell line was U87 (60.4%). Only 10.4% studies used a patient-derived cell line. The proportion of studies not reporting each experimental condition ranged from 8.0–27.4%, including base medium (8.0%), serum supplementation (9.9%) and number of replicates (27.4%). In studies reporting IC50, the median value for U87 at 24 h, 48 h and 72 h was 123.9 μM (IQR 75.3–277.7 μM), 223.1 μM (IQR 92.0–590.1 μM) and 230.0 μM (IQR 34.1–650.0 μM), respectively. The median IC50 at 72 h for patient-derived cell lines was 220 μM (IQR 81.1–800.0 μM). CONCLUSION Temozolomide sensitivity reported in comparable studies was not consistent between or within malignant glioma cell lines. Drug discovery science performed on these models cannot reliably inform clinical translation. A consensus model of reporting can maximise reproducibility and consistency among in vitro studies.

Strategies in synthetic design and structure-activity relationship studies of novel heterocyclic scaffolds as aldose reductase-2 inhibitors.

Heterocyclic scaffolds of natural as well as synthetic origin provide almost all categories of drugs exhibiting a wide range of pharmacological activities, such as antibiotics, antidiabetic and anticancer agents, and so on. Under normal homeostasis, aldose reductase 2 (ALR2) regulates vital metabolic functions; however, in pathological conditions like diabetes, ALR2 is unable to function and leads to secondary diabetic complications. ALR2 inhibitors are a novel target for the treatment of retinopathy (cataract) influenced by diabetes. Epalrestat (stat), an ALR2 inhibitor, is the only drug candidate that was approved in the last four decades; the other drugs from the stat class were retracted after clinical trial studies due to untoward iatrogenic effects. The present studysummarizes the recent development (2014 and onwards) of this pharmacologically active ALR2 heterocyclic scaffold and illustrates the rationale behind the design, structure-activity relationships, and biological studies performed on these molecules. The aim of the current review is to pave a straight path for medicinal chemists and chemical biologists, and, in general, to the drug discovery scientists to facilitate the synthesis and development of novel ALR2 inhibitors that may serve as lead molecules for the treatment of diseases related to the ALR2 enzyme.