Virtual Screening Research Articles

A recombinant L-threonine aldolase with high catalytic efficiency for the asymmetric synthesis of L-threo-phenylserine and L-threo-4-fluorophenylserine.

To develop robust variants of L-threonine aldolases (L-TAs), potent catalysts for synthesizing asymmetric β-hydroxy-α-amino acids, it is necessary to identify critical residues beyond the known active site residues. Through virtual screening, a neglected residue Asn305, was identified as critical for catalytic efficiency. Subsequent site-saturation mutagenesis led to a potent variant N305R which exhibited excellent conversions of 88%conv (87%de) and 80%conv (94%de) for the synthesis of L-threo-phenylserine and L-threo-4-fluorophenylserine respectively. This variant not only outperformed the template enzyme, but also represented a promising L-TA for synthesizing the two β-hydroxy-α-amino acids. It was suggested that Arg305 of the variant N305R generated strong cation-arene interaction and electrostatic force with the intermediates, leading to strengthened binding, enhanced L-threo favored orientation and wider entrance. Our work not only provided an excellent variant N305R, but also suggested the crucial function of a neglected residue Asn305, which offered valuable experiences for other L-TA studies.

Identification of novel tau positron emission tomography tracers for chronic traumatic encephalopathy by comprehensive in silico screening and molecular dynamics simulation.

Chronic traumatic encephalopathy (CTE), a neurodegenerative disease associated with repetitive mild traumatic brain injury, is characterized neuropathologically by abnormal hyperphosphorylated tau accumulation. Early detection of tau deposition in the brain is crucial for the prevention and evaluation of CTE. Positron emission tomography (PET) tracers can image specific proteins, while the optimal PET tracer for CTE tau fibrils remains unidentified. In this study, structure-based virtual screening and CNS PET MPO algorithms were utilized to identify candidates for novel tau PET tracers from 23 000 compounds in the ChemDiv CNS BBB library. A total of 8 μs molecular dynamics simulations were then employed to evaluate their binding affinity and atomic-level interaction with CTE tau protofibrils. The results indicate that V017-7820 (CNS-4), S776-0061 (CNS-12), S567-0465 (CNS-18), and T828-0465 (CNS-25) exhibit higher docking scores and binding free energies with CTE tau protofibrils while also satisfying the fundamental physicochemical properties of PET tracers. Further simulation analyses reveal that CNS-4 has the strongest binding affinity to tau protofibrils among the four compounds. Hydrophobic, π-π stacking, and hydrogen bonding interactions are the primary driving forces for the binding of these compounds to CTE tau protofibrils. In particular, CNS-12 and CNS-25 exhibit more intense hydrophobic and π-π stacking interactions, whereas CNS-4 and CNS-25 exhibit stronger hydrogen bonding interactions. This study identifies promising lead compounds for tau PET tracers and highlights their mechanism of binding to CTE tau protofibrils, which provides new insights for further screening and development of novel PET tracers for CTE diagnosis.

Identify critical genes of breast cancer and corresponding leading natural product compounds of potential therapeutic targets.

Breast cancer is a leading cause of cancer mortality among women globally, with over 2.26 million new cases annually, according to GLOBOCAN 2020. This accounts for approximately 25% of all new female cancers and 15.5% of female cancer deaths.To address this critical public health challenge, we conducted a multi-omics study aimed at identifying hub genes, therapeutic targets, and potential natural product-based therapies. We employed weighted gene co-expression network analysis (WGCNA) and differential gene expression analysis to pinpoint hub genes in breast cancer. Regulatory networks for these genes were constructed by re-analyzing chromatin immunoprecipitation sequencing (ChIP-seq) data from breast cancer cell lines. Additionally, single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics (ST) were utilized to characterize hub gene expression profiles and their relationships with immune cell clusters and tumor microenvironments. Survival analysis based on mRNA and protein expression levels identified prognostic factors and potential therapeutic targets. Lastly, large-scale virtual screening of natural product compounds revealed leading compounds that target squalene epoxidase (SQLE). Our multi-omics analysis paves the way for more effective clinical treatments for breast cancer.

Integration of pharmacophore-based virtual screening, molecular docking, ADMET analysis, and MD simulation for targeting EGFR: A comprehensive drug discovery study using commercial databases.

The epidermal growth factor receptor (EGFR), a crucial component of cellular signaling pathways, is frequently dysregulated in a range of cancers. EGFR targeting has become a viable approach in the development of anti-cancer medications. This study employs an integrated approach to drug discovery, combining multiple computational methodologies to identify potential EGFR inhibitors. The co-crystal ligand for the EGFR protein (R85) (PDB ID: 7AEI) was employed as a model for developing pharmacophore hypotheses. Nine databases underwent a ligand-based virtual screening, and 1271 hits meeting the screening criteria were chosen. EGFR protein crystal structure was obtained from the PDB database (PDB ID: 7AEI) and prepared. The hit compounds identified during virtual screening were docked to the prepared EGFR receptor to predict binding affinities by using the glide tool's standard precision mode. The top ten compounds were chosen, and their affinities of binding ranged from -7.691 to -7.338 kcal/mol. The ADMET properties of the selected compounds were predicted, and three compounds MCULE-6473175764, CSC048452634, and CSC070083626 showed better QPPCaco values compared to other identified compounds, so these were selected for further stability analysis. To confirm the stability of the protein-ligand complexes, a 200 ns molecular dynamics (MD) simulation was run using the binding sites of the top three compounds against the EGFR receptor. These results suggest that the selected compounds may be lead compounds in suppressing the biological activity of EGFR, additional experimental investigation is required.

Abstract A004: Broad and synergistic anti-tumor effects of small-molecule inhibitors of hypoxia inducible factors when paired with immune checkpoint blockade

Abstract Motivation: Hypoxia inducible factors (HIFs) are critical transcriptional regulators that play fundamental roles in adaptive responses to low oxygen, both in normal and disease conditions. Dysregulated activation of HIFs has been implicated in a variety of cancers due to their role in mediating a broad range of key processes including angiogenesis, glycolysis, cell survival, cancer stem cell specification, and metastasis. Methods: In this study, we utilized virtual and cell-based high-throughput screening approaches to identify compounds that disrupt the binding of HIF-1α and HIF-2α to HIF-1β, leading to the suppression of HIF target gene expression. Results: In syngeneic animal models, we observed significant tumor regression when HIF inhibitors are combined with immunotherapy agents in multiple tumor models. The combined treatment enhanced the infiltration of natural killer (NK) cells into the tumor microenvironment. Conclusion: This study highlights the anti-tumor effects of HIF inhibitors and the immune response resulting in sustained tumor regression. The promising outcomes warrant further exploration of these inhibitors as potential cancer treatment agents. Citation Format: Shaima Salman, Tina Y. Huang, Yousang Hwang, Anmol Kumar, Dominic Dordai, Alexander D. MacKerell, Gregg L. Semenza. Broad and synergistic anti-tumor effects of small-molecule inhibitors of hypoxia inducible factors when paired with immune checkpoint blockade [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Optimizing Therapeutic Efficacy and Tolerability through Cancer Chemistry; 2024 Dec 9-11; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Mol Cancer Ther 2024;23(12_Suppl):Abstract nr A004.

Machine learning based algorithms for virtual early detection and screening of neurodegenerative and neurocognitive disorders: a systematic-review

Background and aimNeurodegenerative disorders (e.g., Alzheimer’s, Parkinson’s) lead to neuronal loss; neurocognitive disorders (e.g., delirium, dementia) show cognitive decline. Early detection is crucial for effective management. Machine learning aids in more precise disease identification, potentially transforming healthcare. This comprehensive systematic review discusses how machine learning (ML), can enhance early detection of these disorders, surpassing traditional diagnostics’ constraints.MethodsIn this review, databases were examined up to August 15th, 2023, for ML data on neurodegenerative and neurocognitive diseases using PubMed, Scopus, Google Scholar, and Web of Science. Two investigators used the RAYYAN intelligence tool for systematic reviews to conduct the screening. Six blinded reviewers reviewed titles/abstracts. Cochrane risk of bias tool was used for quality assessment.ResultsOur search found 7,069 research studies, of which 1,365 items were duplicates and thus removed. Four thousand three hundred and thirty four studies were screened, and 108 articles met the criteria for inclusion after preprocessing. Twelve ML algorithms were observed for dementia, showing promise in early detection. Eighteen ML algorithms were identified for Parkinson’s, each effective in detection and diagnosis. Studies emphasized that ML algorithms are necessary for Alzheimer’s to be successful. Fourteen ML algorithms were discovered for mild cognitive impairment, with LASSO logistic regression being the only one with unpromising results.ConclusionThis review emphasizes the pressing necessity of integrating verified digital health resources into conventional medical practice. This integration may signify a new era in the early detection of neurodegenerative and neurocognitive illnesses, potentially changing the course of these conditions for millions globally. This study showcases specific and statistically significant findings to illustrate the progress in the area and the prospective influence of these advancements on the global management of neurocognitive and neurodegenerative illnesses.

Abstract A001: Navigating the future of therapeutic development: innovations in safety and efficacy through advanced research methodologies and nanotechnology

Abstract In recent years, the field of therapeutic development has experienced a significant paradigm shift driven by strategic scientific innovations that prioritize safety and efficacy in therapeutic agents. This paper explores the intersection of advanced research methodologies, novel material sciences, and biotechnological advancements that are revolutionizing the approach to drug formulation and delivery systems. The integration of precision medicine principles alongside cutting-edge technologies allows for the design of therapeutics that are not only effective in targeting specific diseases but are also engineered to minimize adverse effects associated with conventional therapies. One focal point of this exploration is the application of nanotechnology in drug delivery systems. By utilizing nanoparticles, researchers can enhance the bioavailability of therapeutic agents while reducing systemic toxicity. This targeted approach enables the delivery of higher concentrations of drugs to the affected areas, thereby maximizing therapeutic efficacy and minimizing side effects. Additionally, the paper highlights the role of computational modeling and simulations in predicting the interaction of therapeutic agents at the molecular level, facilitating the identification of safer compounds through virtual screening methods. Another significant aspect addressed in this paper is the development of biologics and biosimilars as safer alternatives to traditional synthetic drugs. The advancement of biomanufacturing processes has made it feasible to produce complex molecules with high specificity and reduced immunogenicity, paving the way for innovative therapies that cater to individual patient profiles. Furthermore, the strategic incorporation of pharmacogenomics in drug development allows for the customization of therapeutic agents, ensuring that treatment regimens are tailored to the genetic makeup of patients, which can drastically improve safety and treatment outcomes. The paper also discusses the regulatory landscape surrounding the introduction of these innovative therapeutic agents. It emphasizes the need for adaptive regulatory frameworks that keep pace with scientific advancements while ensuring patient safety and efficacy. Collaborative efforts between researchers, regulatory bodies, and pharmaceutical companies are essential to facilitate the translation of innovative therapies from the laboratory to clinical practice. In conclusion, this paper underscores the critical importance of strategic scientific innovations in the quest for safer therapeutic agents. By harnessing the potential of advanced technologies, researchers can develop therapeutics that not only meet the growing demand for efficacy but also uphold the highest standards of safety, ultimately improving patient care and health outcomes. The findings presented advocate for continued investment in research and collaboration across disciplines to drive future advancements in therapeutic development. Citation Format: Peter O. David. Navigating the future of therapeutic development: innovations in safety and efficacy through advanced research methodologies and nanotechnology. [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Optimizing Therapeutic Efficacy and Tolerability through Cancer Chemistry; 2024 Dec 9-11; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Mol Cancer Ther 2024;23(12_Suppl):Abstract nr A001

Repurposed drugs as PCSK9-LDLR disruptors for lipid lowering and cardiovascular disease therapeutics.

The PCSK9 protein binds to LDL receptors (LDLR), leading to their degradation and reduced expression on cell surfaces. This decreased the clearance of LDL cholesterol from the bloodstream, thereby increasing the risk of coronary artery diseases. Targeting the PCSK9-LDL receptor interaction is crucial for regulating LDL cholesterol levels and preventing cardiovascular disease. This study aims to screen low molecular weight inhibitors to disrupt the PCSK9-LDLR interaction. We employed a comprehensive approach combining high-throughput virtual screening of DrugBank database, followed by molecular docking studies using CDOCKER and flexible docking methods. The top four lead compounds were further validated through molecular dynamics (MD) simulations and binding free energy calculations using MM-PBSA. Finally, the in vitro assay confirmed that Benazepril and Quinapril exhibited the highest potency as PCSK9-LDLR disruptors among the top candidates. These lead compounds have the potential to be repurposed as lipid-lowering agents for the treatment of cardiovascular diseases, offering a promising therapeutic strategy.

IDCNNPred: an interpretable deep learning model for virtual screening and identification of PI3Ka inhibitors against triple-negative breast cancer.

Triple-negative breast cancer (TNBC) lacks estrogen, progesterone, and HER2 expression, accounting for 15-20% of breast cancer cases. It is challenging due to low therapeutic response, heterogeneity, and aggressiveness. The PI3Ka isoform is a promising therapeutic target, often hyperactivated in TNBC, contributing to uncontrolled growth and cancer cell formation. We have proposed an interpretable deep convolutional neural network prediction (iDCNNPred) system using 2D molecular images to classify bioactivity and identify potential PI3Ka inhibitors. We built Custom-DCNN models and pre-trained models such as AlexNet, SqueezeNet, and VGG19 by using the Bayesian optimization algorithm, and found that our Custom-DCNN model performed better than a pre-trained model with lower complexity and memory usage. All top-performed models were screened with the Maybridge Chemical library to find predictive hit molecules. The screened molecules were further evaluated for protein-ligand interaction with molecular docking and finally 12 promising hits were shortlisted for biological validation using in-vitro PI3K inhibition studies. After biological evaluation, 4 potent molecules with different structural moieties were identified, and these molecules present new starting scaffolds for further improvement in terms of their potency and selectivity as PI3K inhibitors with the help of medicinal chemistry efforts. Furthermore, we also showed the significance of the interpretation and visualization of the model's predictions by the Grad-CAM technique, enhancing the robustness, transparency, and interpretability of the model's predictions. The data and script files and prediction run of models used for this study to reproduce the experiment are available in the GitHub repository at https://github.com/ravishankar1307/iDCNNPred.git .

Discovery of Novel MyD88 Inhibitor A5S to Alleviate Acute Lung Injury with Favorable Drug-like Properties.

Myeloid differentiation primary response 88 (MyD88) plays a central role in inflammatory responses and diseases. However, only a few inhibitors of MyD88 with some limits have been reported currently. Herein, we identified a lead compound (L7) through virtual screening and synthesized twenty-seven L7 derivatives. An optimal compound (A5) was determined through enzyme-linked immunosorbent assay (ELISA), 2,5-diphenyl-2H-tetrazolium bromide (MTT), and biolayer interferometry (BLI) assay. The potent isomer A5S showed a high MyD88 binding ability and exerted an anti-inflammatory effect through the NF-κB/MAPK pathway. A5S had good stability and safety, showed the highest distribution concentration in the lungs, and exhibited good therapeutic effects on LPS-induced and sepsis-induced ALI mouse models. Most importantly, A5S showed advantages in PK properties, and was identified as a promising MyD88 inhibitor with favorable drug-like properties, compared to the only approved MyD88 inhibitor, TJ-M2010-5, which is currently undergoing a Phase I study, and our previously reported MyD88 inhibitors LM8.

Structure-based Virtual Screening and Biological Characterization of Novel BACE-1 and Amyloid-β Aggregation Inhibitors.

Alzheimer's disease (AD) is a complex neurodegenerative disorder having limited treatment options. The beta-site APP cleaving enzyme 1 (BACE-1) is a key target for therapeutic intervention in Alzheimer's disease. To discover new scaffolds for BACE-1 inhibitors, a ChemBridge DIVERSet library of 20,000 small molecules was employed to structure-based virtual screening. The top 45 compounds, based on docking scores and binding affinities, were tested for BACE-1 inhibitory activity using a FRET assay. Four compounds, 18 (5353320), 20 (5262831), 29 (5784196) and 32 (5794006) demonstrated more than 35% inhibitory activity at 10 µM. Notably, pyrazole-5-carbohydrazide 29 (5784196) exhibited BACE-1 inhibition with an IC50 value of 14.5 µM and a ki value of 0.25 μM. Additionally, it also inhibits the self-aggregation of β-amyloid, with IC50 value of 14.87 µM. Molecular modeling and dynamics simulations provided insights into its interaction pattern and stability of the enzyme-inhibitor complex. These findings suggest that virtual screening is an efficient and cost-effective method for identifying potential leads for AD.

Identification of potential methyltransferase NSD2 enzymatic inhibitors through a multi-step structure-based drug design.

Reversing aberrant protein methylation levels is widely recognized as a key focus in cancer therapy. As an essential lysine methylation regulator, NSD2 (Nuclear receptor-binding SET Domain 2, also known as WHSC1/MMSET) regulates chromatin structural sparsity and DNA repair processes. Abnormal enhancement of NSD2 methylation activity (caused by NSD2 overexpression and point mutations) has been closely related to the initiation and development of various cancers and diseases. However, the lack of selective inhibitors hinders further therapeutic intervention and limits the exploration of its biological mechanism. Therefore, this study developed an integrated approach that includes binding feature pharmacophore modeling, gradient database screening of 120 million compounds, flexible docking, and molecular dynamic simulation. This approach was used to identify hit compounds targeting the substrate/coenzyme binding site of NSD2. Subsequently, 20 lead compounds were retrieved by using molecular docking analysis and ADMET prediction. Finally, MD simulations were performed to validate the binding stability of selected drug candidates. The findings indicated that these newly obtained compounds might be potent NSD2 inhibitors. We hope the integrated virtual screening approach will provide a valuable idea for discovering novel H3K36 methyltransferase inhibitors.

Analog Accessibility Score (AAscore) for Rational Compound Selection.

Various in silico scores have been proposed to objectively assess the characteristics and properties of a compound. However, there is still no score that represents the analog accessibility of a compound. Such a score would be valuable for selecting compounds proposed by virtual screening or for prioritizing hit compounds for the hit-to-lead phase. This study proposes an analog accessibility score (AAscore), where retrosynthesis prediction and forward product prediction models were utilized to generate virtual analogs. The AAscore is defined as the number of unique analogs and virtual synthetic routes. To evaluate the AAscore in terms of the number of actually synthesized analog compounds, analog compounds were prepared by using the compound-core relationship (CCR) method. It was found that the AAscore was little correlated with the number of CCR-based analogs. Furthermore, AAscores were found to be significantly influenced by the number of extracted candidate reactants from a reactant database. A case study targeting compounds active against carbonic anhydrase 2 showed that the AAscore could identify compounds that were synthesized into analogs.

Identification of Novel Drug Molecules Against NS3-Like Helicase Enzyme of Alongshan Virus.

Alongshan virus (ALSV) is a novel tick-borne virus associated with human diseases. The ALSV is a segmented flavivirus from the family Flaviviridae. It is currently considered as tick-borne arbovirus. There is a high incidence of fever and headache among patients with ALSV infection, and some patients also present with fatigue, coma, depression, nausea, myalgia/arthralgia, and skin rashes. Neither a licensed vaccine nor a drug is currently available to treat ALSV. The development of new, practical, and innovative therapeutic approaches is needed to overcome the emergence of the pathogen. Research on drugs remains a complex, time-consuming, and expensive. The field of drug development has undergone a revolution due to the use ofcomputational approaches, which provide several benefits that speed up and improve the process of developing novel drugs. The goal of this study is to identify novel drug-like molecules against NS3-like helicase enzyme of Alongshan virus. Using molecular docking, the binding potential of the top three ligands to the specified targetwas determined. Molecular dynamic simulations wereused to identifythe stabilities of the best-dockedconformations followed by energy calculations and ADMET analysis. Three potential and promising compounds were identified by performing structure-based virtual screening of non-structural protein 3 (NS3) like helicase of Alongshan virus. The best-docked complexes identified through virtual screening were BDC-23169381, BDB-26412846, BDB-2641954. All these compounds had good pharmacokinetics characteristics and were identified as drug like.

Integrative Machine Learning, Virtual Screening, and Molecular Modeling for BacA-Targeted Anti-Biofilm Drug Discovery Against Staphylococcal Infections

The rise in antibiotic-resistant Staphylococcal infections necessitates innovative approaches to identify new therapeutic agents. This study investigates the application of machine learning models to identify potential phytochemical inhibitors against BacA, a target related to Staphylococcal infections. Active compounds were retrieved from BindingDB while the decoy was generated from DUDE. The RDKit was utilized for feature engineering. Machine learning models such as k-nearest neighbors (KNN), the support vector machine (SVM), random forest (RF), and naive Bayes (NB) were trained on an initial dataset consisting of 226 active chemicals and 2550 inert compounds. Accompanied by an MCC of 0.93 and an accuracy of 96%, the RF performed better. Utilizing the RF model, a library of 9000 phytochemicals was screened, identifying 300 potentially active compounds, of which 192 exhibited drug-like properties and were further analyzed through molecular docking studies. Molecular docking results identified Ergotamine, Withanolide E, and DOPPA as top inhibitors of the BacA protein, accompanied by interaction affinities of −8.8, −8.1, and −7.9 kcal/mol, respectively. Molecular dynamics (MD) was applied for 100 ns to these top hits to evaluate their stability and dynamic behavior. RMSD, RMSF, SASA, and Rg analyses showed that all complexes remained stable throughout the simulation period. Binding energy calculations using MMGBSA analysis revealed that the BacA_Withanolide E complex exhibited the most favorable binding energy profile with significant van der Waals interactions and a substantial reduction in gas-phase energy. It also revealed that van der Waals interactions contributed significantly to the binding stability of Withanolide E, while electrostatic interactions played a secondary role. The integration of machine learning models with molecular docking and MD simulations proved effective in identifying promising phytochemical inhibitors, with Withanolide E emerging as a potent candidate. These findings provide a pathway for developing new antibacterial agents against Staphylococcal infections, pending further experimental validation and optimization.

Ultra-Large Virtual Screening: Definition, Recent Advances, and Challenges in Drug Design.

Virtual screening (VS) in drug design employs computational methodologies to systematically rank molecules from a virtual compound library based on predicted features related to their biological activities or chemical properties. The recent expansion in commercially accessible compound libraries and the advancements in artificial intelligence (AI) and computational power - including enhanced central processing units (CPUs), graphics processing units (GPUs), high-performance computing (HPC), and cloud computing - have significantly expanded our capacity to screen libraries containing over 109 molecules. Herein, we review the concept of ultra-large virtual screening (ULVS), focusing on the various algorithms and methodologies employed for virtual screening at this scale. In this context, we present the software utilized, applications, and results of different approaches, such as brute force docking, reaction-based docking approaches, machine learning (ML) strategies applied to docking or other VS methods, and similarity/pharmacophore search-based techniques. These examples represent a paradigm shift in the drug discovery process, demonstrating not only the feasibility of billion-scale compound screening but also their potential to identify hit candidates and increase the structural diversity of novel compounds with biological activities.

Furopyridine Derivatives as Potent Inhibitors of the Wild Type, L858R/T790M, and L858R/T790M/C797S EGFR.

The treatment of patients with nonsmall cell lung cancer (NSCLC) using epidermal growth factor receptor (EGFR) inhibitors is complicated by drug-sensitive activating L858R/T790M and L858R/T790M/C797S mutations. To overcome drug resistance, a series of furopyridine (PD) compounds were virtually screened to identify potent EGFR inhibitors using molecular docking and molecular dynamics (MD) simulations based on the solvated interaction energy (SIE) method. Several PD compounds identified from virtual screening demonstrated the potential to suppress both wild-type and mutant forms of EGFR, with IC50 values in the nanomolar range. Among these, PD18 and PD56 exhibited highly potent inhibitory activity against both wild-type and mutant forms of EGFR, surpassing the efficacy of known drugs. Additionally, both PD compounds were cytotoxic to NSCLC cell lines (A549 and H1975) while being nontoxic to normal cell lines (Vero). The interaction mechanisms of both PD compounds complexed with wild-type and mutant forms of EGFR were elucidated through 500 ns molecular dynamics simulations. The predicted binding affinity from molecular mechanics/Poisson-Boltzmann surface area (MM/PBSA) correlated well with the experimental binding affinity derived from IC50 values. Furthermore, it was observed that van der Waals interactions, rather than electrostatic interactions, played a significant role in interacting with EGFR's active site. The strong inhibitory activity against EGFR was attributed to two key residues, M793 and S797, via hydrogen bonding, corresponding with lower solvent accessibility and a higher number of atomic contacts. Therefore, these potent compounds could be developed as promising drugs targeting both wild-type and mutant EGFR for the treatment of NSCLC.

CSearch: chemical space search via virtual synthesis and global optimization

The two key components of computational molecular design are virtually generating molecules and predicting the properties of these generated molecules. This study focuses on an effective method for molecular generation through virtual synthesis and global optimization of a given objective function. Using a pre-trained graph neural network (GNN) objective function to approximate the docking energies of compounds for four target receptors, we generated highly optimized compounds with 300–400 times less computational effort compared to virtual compound library screening. These optimized compounds exhibit similar synthesizability and diversity to known binders with high potency and are notably novel compared to library chemicals or known ligands. This method, called CSearch, can be effectively utilized to generate chemicals optimized for a given objective function. With the GNN function approximating docking energies, CSearch generated molecules with predicted binding poses to the target receptors similar to known inhibitors, demonstrating its effectiveness in producing drug-like binders.Scientific Contribution We have developed a method for effectively exploring the chemical space of drug-like molecules using a global optimization algorithm with fragment-based virtual synthesis. The compounds generated using this method optimize the given objective function efficiently and are synthesizable like commercial library compounds. Furthermore, they are diverse, novel drug-like molecules with properties similar to known inhibitors for target receptors.

Machine learning-aided discovery of T790M-mutant EGFR inhibitor CDDO-Me effectively suppresses non-small cell lung cancer growth

BackgroundEpidermal growth factor receptor (EGFR) T790M mutation often occurs during long durational erlotinib treatment of non-small cell lung cancer (NSCLC) patients, leading to drug resistance and disease progression. Identification of new selective EGFR-T790M inhibitors has proven challenging through traditional screening platforms. With great advances in computer algorithms, machine learning improved the screening rates of molecules at full chemical spaces, and these molecules will present higher biological activity and targeting efficiency.MethodsAn integrated machine learning approach, integrated by Bayesian inference, was employed to screen a commercial dataset of 70,413 molecules, identifying candidates that selectively and efficiently bind with EGFR harboring T790M mutation. In vitro cellular assays and molecular dynamic simulations was used for validation. EGFR knockout cell line was generated for cross-validation. In vivo xenograft moues model was constructed to investigate the antitumor efficacy of CDDO-Me.ResultsOur virtual screening and subsequent in vitro testing successfully identified CDDO-Me, an oleanolic acid derivative with anti-inflammatory activity, as a potent inhibitor of NSCLC cancer cells harboring the EGFR-T790M mutation. Cellular thermal shift assay and molecular dynamic simulation validated the selective binding of CDDO-Me to T790M-mutant EGFR. Further experimental results revealed that CDDO-Me induced cellular apoptosis and caused cell cycle arrest through inhibiting the PI3K-Akt-mTOR axis by directly targeting EGFR protein, cross-validated by sgEGFR silencing in H1975 cells. Additionally, CDDO-Me could dose-depended suppress the tumor growth in a H1975 xenograft mouse model.ConclusionCDDO-Me induced apoptosis and caused cell cycle arrest by inhibiting the PI3K-Akt-mTOR pathway, directly targeting the EGFR protein. In vivo studies in a H1975 xenograft mouse model demonstrated dose-dependent suppression of tumor growth. Our work highlights the application of machine learning-aided drug screening and provides a promising lead compound to conquer the drug resistance of NSCLC.Graphical

Probing the Conformational Ensemble of the Amyloid Beta 16-22 Fragment with Parallel-Bias Metadynamics.

Aβ(16-22) is a segment of the Alzheimer's-related β-amyloid peptide that plays a crucial role in its aggregation. This study applies well-tempered parallel-bias metadynamics to investigate the impact of several denaturants and osmolytes on the conformational ensembles of both termini-capped and uncapped Aβ(16-22) monomers. Comparison of the different sets of collective variables in the metadynamics bias shows that using the set of backbone torsional angles results in better and faster convergence of simulations than employing more general structural characteristics of the short peptide. The equilibrium conformational ensembles of the peptides are characterized in pure water and in the presence of TMAO, urea, guanidinium chloride, and trifluoroethanol. In particular, trifluoroethanol and TMAO are found to increase the population of compact peptide conformations, whereas urea and guanidinium chloride favor extended structures. The analysis of the free energy surfaces in the presence of various substances with a comparison of the behavior of the capped and uncapped peptide forms reveals the role of different types of intrapeptide interactions such as salt bridges, hydrophobic contacts, and hydrogen bonds in stabilization of the compact or extended structures. As compounds reducing the abundance of the compact states of Aβ(16-22) and other disordered peptides are likely to suppress their amyloid fibril formation, simulations in the systems with this short peptide may be useful for the virtual screening of such compounds.