Search

Artificial intelligence transforms drug discovery, with phenotype-based approaches emerging as a promising alternative to target-based methods, overcoming limitations like lack of well-defined targets. While chemical-induced transcriptional profiles offer a comprehensive view of drug mechanisms, inherent noise often obscures the true signal, hindering their potential for meaningful insights. Here, we highlight the development of TranSiGen, a deep generative model employing self-supervised representation learning. TranSiGen analyzes basal cell gene expression and molecular structures to reconstruct chemical-induced transcriptional profiles with high accuracy. By capturing both cellular and compound information, TranSiGen-derived representations demonstrate efficacy in diverse downstream tasks like ligand-based virtual screening, drug response prediction, and phenotype-based drug repurposing. Notably, in vitro validation of TranSiGen's application in pancreatic cancer drug discovery highlights its potential for identifying effective compounds. We envisage that integrating TranSiGen into the drug discovery and mechanism research holds significant promise for advancing biomedicine. While chemical-induced transcriptional profiles reveal drug mechanisms, inherent noise limits their utility. Here, authors present TranSiGen, a deep representation learning model that denoises and reconstructs these profiles, demonstrating its efficacy in downstream drug discovery tasks. [ABSTRACT FROM AUTHOR]

As the most abundant organic substances in nature, carbohydrates are essential for life. Understanding how carbohydrates regulate proteins in the physiological and pathological processes presents opportunities to address crucial biological problems and develop new therapeutics. However, the diversity and complexity of carbohydrates pose a challenge in experimentally identifying the sites where carbohydrates bind to and act on proteins. Here, we introduce a deep learning model, DeepGlycanSite, capable of accurately predicting carbohydrate-binding sites on a given protein structure. Incorporating geometric and evolutionary features of proteins into a deep equivariant graph neural network with the transformer architecture, DeepGlycanSite remarkably outperforms previous state-of-the-art methods and effectively predicts binding sites for diverse carbohydrates. Integrating with a mutagenesis study, DeepGlycanSite reveals the guanosine-5'-diphosphate-sugar-recognition site of an important G-protein coupled receptor. These findings demonstrate DeepGlycanSite is invaluable for carbohydrate-binding site prediction and could provide insights into molecular mechanisms underlying carbohydrate-regulation of therapeutically important proteins. Carbohydrates are essential for regulating various biological processes. Here, the authors developed DeepGlycanSite, a deep learning model that accurately predicts carbohydrate-binding sites on proteins, offering insights into carbohydrate regulation of therapeutically important proteins. [ABSTRACT FROM AUTHOR]

Lipophilicity is a fundamental physical property that significantly affects various aspects of drug behavior, including solubility, permeability, metabolism, distribution, protein binding, and toxicity. Accurate prediction of lipophilicity, measured by the logD7.4 value (the distribution coefficient between n-octanol and buffer at physiological pH 7.4), is crucial for successful drug discovery and design. However, the limited availability of data for logD modeling poses a significant challenge to achieving satisfactory generalization capability. To address this challenge, we have developed a novel logD7.4 prediction model called RTlogD, which leverages knowledge from multiple sources. RTlogD combines pre-training on a chromatographic retention time (RT) dataset since the RT is influenced by lipophilicity. Additionally, microscopic pKa values are incorporated as atomic features, providing valuable insights into ionizable sites and ionization capacity. Furthermore, logP is integrated as an auxiliary task within a multitask learning framework. We conducted ablation studies and presented a detailed analysis, showcasing the effectiveness and interpretability of RT, pKa, and logP in the RTlogD model. Notably, our RTlogD model demonstrated superior performance compared to commonly used algorithms and prediction tools. These results underscore the potential of the RTlogD model to improve the accuracy and generalization of logD prediction in drug discovery and design. In summary, the RTlogD model addresses the challenge of limited data availability in logD modeling by leveraging knowledge from RT, microscopic pKa, and logP. Incorporating these factors enhances the predictive capabilities of our model, and it holds promise for real-world applications in drug discovery and design scenarios. [ABSTRACT FROM AUTHOR]

Three-dimensional (3D) conformations of a small molecule profoundly affect its binding to the target of interest, the resulting biological effects, and its disposition in living organisms, but it is challenging to accurately characterize the conformational ensemble experimentally. Here, we proposed an autoregressive torsion angle prediction model Tora3D for molecular 3D conformer generation. Rather than directly predicting the conformations in an end-to-end way, Tora3D predicts a set of torsion angles of rotatable bonds by an interpretable autoregressive method and reconstructs the 3D conformations from them, which keeps structural validity during reconstruction. Another advancement of our method over other conformational generation methods is the ability to use energy to guide the conformation generation. In addition, we propose a new message-passing mechanism that applies the Transformer to the graph to solve the difficulty of remote message passing. Tora3D shows superior performance to prior computational models in the trade-off between accuracy and efficiency, and ensures conformational validity, accuracy, and diversity in an interpretable way. Overall, Tora3D can be used for the quick generation of diverse molecular conformations and 3D-based molecular representation, contributing to a wide range of downstream drug design tasks. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI)-based molecular design methods, especially deep generative models for generating novel molecule structures, have gratified our imagination to explore unknown chemical space without relying on brute-force exploration. However, whether designed by AI or human experts, the molecules need to be accessibly synthesized and biologically evaluated, and the trial-and-error process remains a resources-intensive endeavor. Therefore, AI-based drug design methods face a major challenge of how to prioritize the molecular structures with potential for subsequent drug development. This study indicates that common filtering approaches based on traditional screening metrics fail to differentiate AI-designed molecules. To address this issue, we propose a novel molecular filtering method, MolFilterGAN, based on a progressively augmented generative adversarial network. Comparative analysis shows that MolFilterGAN outperforms conventional screening approaches based on drug-likeness or synthetic ability metrics. Retrospective analysis of AI-designed discoidin domain receptor 1 (DDR1) inhibitors shows that MolFilterGAN significantly increases the efficiency of molecular triaging. Further evaluation of MolFilterGAN on eight external ligand sets suggests that MolFilterGAN is useful in triaging or enriching bioactive compounds across a wide range of target types. These results highlighted the importance of MolFilterGAN in evaluating molecules integrally and further accelerating molecular discovery especially combined with advanced AI generative models. [ABSTRACT FROM AUTHOR]

CD47 is a macrophage-specific immune checkpoint protein acting by inhibiting phagocytosis. However, the underlying mechanism maintaining CD47 protein stability in cancer is not clear. Here we show that CD47 undergoes degradation via endocytosis/lysosome pathway. The lysosome protein RAGA interacts with and promotes CD47 lysosome localization and degradation. Disruption of RAGA blocks CD47 degradation, leading to CD47 accumulation, high plasma membrane/intracellular CD47 expression ratio and reduced phagocytic clearance of cancer cells. RAGA deficiency promotes tumor growth due to the accumulation of CD47, which sensitizes the tumor to CD47 blockade. Clinical analysis shows that RAGA and CD47 proteins are negatively correlated in lung adenocarcinoma patient samples. High RAGA protein level is related to longer patient survival. In addition, RAGAhighCD47low patients show the longest overall survival. Our study thereby not only reveals a mechanism by which RAGA regulates CD47 lysosome degradation, but also suggests RAGA is a potential diagnostic biomarker of lung adenocarcinoma. The lysosome protein RAGA interacts with and promotes lysosome localization and degradation of the macrophage-specific immune checkpoint protein CD47, with possible relevance to lung adenocarcinoma [ABSTRACT FROM AUTHOR]

PROteolysis TArgeting Chimeras (PROTACs) has been exploited to degrade putative protein targets. However, the antitumor performance of PROTACs is impaired by their insufficient tumour distribution. Herein, we present de novo designed polymeric PROTAC (POLY-PROTAC) nanotherapeutics for tumour-specific protein degradation. The POLY-PROTACs are engineered by covalently grafting small molecular PROTACs onto the backbone of an amphiphilic diblock copolymer via the disulfide bonds. The POLY-PROTACs self-assemble into micellar nanoparticles and sequentially respond to extracellular matrix metalloproteinase-2, intracellular acidic and reductive tumour microenvironment. The POLY-PROTAC NPs are further functionalized with azide groups for bioorthogonal click reaction-amplified PROTAC delivery to the tumour tissue. For proof-of-concept, we demonstrate that tumour-specific BRD4 degradation with the bioorthogonal POLY-PROTAC nanoplatform combine with photodynamic therapy efficiently regress tumour xenografts in a mouse model of MDA-MB-231 breast cancer. This study suggests the potential of the POLY-PROTACs for precise protein degradation and PROTAC-based cancer therapy. Proteolysis targeting chimeras (PROTACs) have emerged as promising cancer therapy agents but have suffered from systemic toxicity issues. Here, the authors report on the creation of polymeric PROTAC nanoparticles for tumour targeting delivery and demonstrate protein degradation in vivo, in combination with photodynamic therapy. [ABSTRACT FROM AUTHOR]

Blood–brain barrier is a pivotal factor to be considered in the process of central nervous system (CNS) drug development, and it is of great significance to rapidly explore the blood–brain barrier permeability (BBBp) of compounds in silico in early drug discovery process. Here, we focus on whether and how uncertainty estimation methods improve in silico BBBp models. We briefly surveyed the current state of in silico BBBp prediction and uncertainty estimation methods of deep learning models, and curated an independent dataset to determine the reliability of the state-of-the-art algorithms. The results exhibit that, despite the comparable performance on BBBp prediction between graph neural networks-based deep learning models and conventional physicochemical-based machine learning models, the GROVER-BBBp model shows greatly improvement when using uncertainty estimations. In particular, the strategy combined Entropy and MC-dropout can increase the accuracy of distinguishing BBB + from BBB − to above 99% by extracting predictions with high confidence level (uncertainty score < 0.1). Case studies on preclinical/clinical drugs for Alzheimer' s disease and marketed antitumor drugs that verified by literature proved the application value of uncertainty estimation enhanced BBBp prediction model, that may facilitate the drug discovery in the field of CNS diseases and metastatic brain tumors. [ABSTRACT FROM AUTHOR]

The interaction between Lymphocyte function-associated antigen 1 (LFA-1) and intercellular-adhesion molecule-1 (ICAM-1) plays important roles in the cell-mediated immune response and inflammation associated with dry eye disease. LFA-1/ICAM-1 antagonists can be used for the treatment of dry eye disease, such as Lifitegrast which has been approved by the FDA in 2016 as a new drug for the treatment of dry eye disease. In this study, we designed and synthesized some new structure compounds that are analogues to Lifitegrast, and their biological activities were evaluated by in vitro cell-based assay and also by in vivo mouse dry eye model. Our results demonstrated that one of these analogues of Lifitegrast (compound 1b) showed good LFA-1/ICAM-1 antagonist activity in in vitro assay; meanwhile, it also significantly reduced ocular surface epithelial cells damage, increased goblet cell density in dry eye mouse and highly improved the symptoms of dry eye mouse. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) models usually require large amounts of high-quality training data, which is in striking contrast to the situation of small and biased data faced by current drug discovery pipelines. The concept of federated learning has been proposed to utilize distributed data from different sources without leaking sensitive information of the data. This emerging decentralized machine learning paradigm is expected to dramatically improve the success rate of AI-powered drug discovery. Here, we simulated the federated learning process with different property and activity datasets from different sources, among which overlapping molecules with high or low biases exist in the recorded values. Beyond the benefit of gaining more data, we also demonstrated that federated training has a regularization effect superior to centralized training on the pooled datasets with high biases. Moreover, different network architectures for clients and aggregation algorithms for coordinators have been compared on the performance of federated learning, where personalized federated learning shows promising results. Our work demonstrates the applicability of federated learning in predicting drug-related properties and highlights its promising role in addressing the small and biased data dilemma in drug discovery. [ABSTRACT FROM AUTHOR]

The planarian Dugesia japonica has a powerful regenerative capacity and can regenerate its whole body, including complete and functional restoration of its central nervous system (CNS) within one week. In our previous works, down-regulation of DjPLCβ led to defects in brain regeneration that resulted in incomplete brain development, suppression of DjGAD protein expression, and reduction of GABAergic neurons. U73122 is an aminosteroid inhibitor of phospholipase C (PLC), which plays an inhibitory role in many different types of cells, such as platelets, neutrophils, and neuroblastoma cells. m-3M3FBS is an activator that stimulates PLC activity. This study aimed to investigate whether PLCβ could be used as a drug target for GABAergic neuron regeneration. Herein, planarianns were treated with 0.2 μM U73122 or alternatively first treated with 0.2 μM U73122 and then with 1 μM m-3M3FBS for 1 h, and the expression pattern of DjGAD and DjPLCβ in the brain was analyzed by immunohistochemistry; level of DjGAD mRNA and DjPLCβ protein were analyzed by whole-mount in situ hybridization (WISH) and western blotting, respectively. The results showed that U73122 can strongly inhibit DjPLCβ and DjGAD protein expression and cause neuronal defects in planarian. m-3M3FBS rescued the expression of DjPLCβ and DjGAD proteins and mitigated the associated neuronal defects. These results suggest that U73122 and m-3M3FBS may regulate GABAergic neuron regeneration in planarian, and that PLCβ can be used as a novel drug target for GABAergic neuron regeneration following disease-induced neuron loss. [ABSTRACT FROM AUTHOR]

Objective: This study aimed to investigate the inhibitory effect of berberine hydrochloride (BH) on Candida albicans (C.albicans) ATCC10231 biofilm formation. Results: This paper found a positive correlation between the concentration of BH and its inhibitory effect on the cellular activity of early biofilms because we found that 128 and 32 μg/mL BH significantly inhibited biofilm formation (P < 0.05). BH significantly inhibited the cellular activity in early biofilms, destroyed the microscopic morphology of C.albicans and reduced the thickness of the biofilm. Both 128 and 32 μg/mL concentration solutions of BH significantly inhibited biofilm formation (P < 0.05). We found that the inhibitory effect of BH solution was positively correlated with its concentration and 128 μg/mL BH was better than 4 μg/mL fluconazole. Additionally, the results of RT-PCR indicated that 128 and 32 μg/mL BH inhibited the expression of EFG1, HWP1, ECE1, and ALS1 (P < 0.05). Conclusion: The efficacy of BH in inhibiting the formation of C.albicans biofilm by killing the cells in the biofilm and destroying its structure; and the mechanism may be to down-regulate the expression of EFG1, HWP1, ECE1, and ALS1 in hyphae formation, thereby, retarding the morphological transformation of C. albicans. [ABSTRACT FROM AUTHOR]

Thanks to the fast improvement of the computing power and the rapid development of the computational chemistry and biology, the computer-aided drug design techniques have been successfully applied in almost every stage of the drug discovery and development pipeline to speed up the process of research and reduce the cost and risk related to preclinical and clinical trials. Owing to the development of machine learning theory and the accumulation of pharmacological data, the artificial intelligence (AI) technology, as a powerful data mining tool, has cut a figure in various fields of the drug design, such as virtual screening, activity scoring, quantitative structure-activity relationship (QSAR) analysis, de novo drug design, and in silico evaluation of absorption, distribution, metabolism, excretion and toxicity (ADME/T) properties. Although it is still challenging to provide a physical explanation of the AI-based models, it indeed has been acting as a great power to help manipulating the drug discovery through the versatile frameworks. Recently, due to the strong generalization ability and powerful feature extraction capability, deep learning methods have been employed in predicting the molecular properties as well as generating the desired molecules, which will further promote the application of AI technologies in the field of drug design. [ABSTRACT FROM AUTHOR]

Background: Experimental approaches for determining the metabolic properties of the drug candidates are usually expensive, time-consuming and labor intensive. There is a great deal of interest in developing computational methods to accurately and efficiently predict the metabolic decomposition of drug-like molecules, which can provide decisive support and guidance for experimentalists. Results: Here, we developed an integrated, low false positive and reaction types extensive metabolism prediction approach called RD-Metabolizer (Reaction Database-based Metabolizer). RD-Metabolizer firstly employed the detailed reaction SMARTS patterns to encode different metabolism reaction types with the aim of covering larger chemical reaction space. 2D fingerprint similarity calculation model was built to calculate the metabolic probability of each site in a molecule. RDKit was utilized to act on pre-written reaction SMARTS patterns to correct the metabolic ranking of each site in a molecule generated by the 2D fingerprint similarity calculation model as well as generate corresponding structures of metabolites, thus helping to reduce the false positive metabolites. Two test sets were adopted to evaluate the performance of RD-Metabolizer in predicting SOMs and structures of metabolites. The results indicated that RD-Metabolizer was better than or at least as good as several widely used SOMs prediction methods. Besides, the number of false positive metabolites was obviously reduced compared with MetaPrint2D-React. Conclusions: The accuracy and efficiency of RD-Metabolizer was further illustrated by a metabolism prediction case of AZD9291, which is a mutant-selective EGFR inhibitor. RD-Metabolizer will serve as a useful toolkit for the early metabolic properties assessment of drug-like molecules at the preclinical stage of drug discovery. [ABSTRACT FROM AUTHOR]

Compound-protein interactions play important roles in every cell via the recognition and regulation of specific functional proteins. The correct identification of compound-protein interactions can lead to a good comprehension of this complicated system and provide useful input for the investigation of various attributes of compounds and proteins. In this study, we attempted to understand this system by extracting properties from both proteins and compounds, in which proteins were represented by gene ontology and KEGG pathway enrichment scores and compounds were represented by molecular fragments. Advanced feature selection methods, including minimum redundancy maximum relevance, incremental feature selection, and the basic machine learning algorithm random forest, were used to analyze these properties and extract core factors for the determination of actual compound-protein interactions. Compound-protein interactions reported in The Binding Databases were used as positive samples. To improve the reliability of the results, the analytic procedure was executed five times using different negative samples. Simultaneously, five optimal prediction methods based on a random forest and yielding maximum MCCs of approximately 77.55 % were constructed and may be useful tools for the prediction of compound-protein interactions. This work provides new clues to understanding the system of compound-protein interactions by analyzing extracted core features. Our results indicate that compound-protein interactions are related to biological processes involving immune, developmental and hormone-associated pathways. [ABSTRACT FROM AUTHOR]

Correction to: I Nature Communications i https://doi.org/10.1038/s41467-022-32050-4, published online 26 July 2022 The original version of this Article contained an error in Figures 6k (PED + N3@PGDA7 + Laser) and 6l (PED + N3@PGDA7), in which the images were duplicates of adjacent images and the correct images were missing. The original article can be found online at https://doi.org/10.1038/s41467-022-32050-4. [Extracted from the article]

Methanosaeta strains are frequently involved in the granule formation during methanogenic wastewater treatment. To investigate the impact of Methanosaeta on granulation and performance of upflow anaerobic sludge blanket (UASB) reactors, three 1-L working volume reactors noted as R1, R2, and R3 were operated fed with a synthetic wastewater containing sodium acetate and glucose. R1 was inoculated with 1-L activated sludge, while R2 and R3 were inoculated with 200-mL concentrated pre-grown Methanosaeta harundinacea 6Ac culture and 800 mL of activated sludge. Additionally, R3 was daily dosed with 0.5 mL/L of acetyl ether extract of 6Ac spent culture containing its quorum sensing signal carboxyl acyl homoserine lactone (AHL). Compared to R1, R2 and R3 had a higher and more constant chemical oxygen demand (COD) removal efficiency and alkaline pH (8.2) during the granulation phase, particularly, R3 maintained approximately 90 % COD removal. Moreover, R3 formed the best granules, and microscopic images showed fluorescent Methanosaeta-like filaments dominating in the R3 granules, but rod cells dominating in the R2 granules. Analysis of 16S rRNA gene libraries showed increased diversity of methanogen species like Methanosarcina and Methanospirillum in R2 and R3, and increased bacteria diversity in R3 that included the syntrophic propionate degrader Syntrophobacter. Quantitative PCR determined that 6Ac made up more than 22 % of the total prokaryotes in R3, but only 3.6 % in R2. The carboxyl AHL was detected in R3. This work indicates that AHL-facilitated filaments of Methanosaeta contribute to the granulation and performance of UASB reactors, likely through immobilizing other functional microorganisms. [ABSTRACT FROM AUTHOR]

The effect of pH (4.0, 5.0, and 6.0) on acidification and fermentation of fruit and vegetable wastes was investigated using batch and semi-continuous experiments under mesophilic condition. Results showed that fermentation types change with pH variation. The pH of acidification system containing fruit and vegetable wastes could automatically decrease to 3.0 ~ 4.0. At this pH range, a stable ethanol production was observed, at which ethanol-type fermentation was obtained. Based on the results, the fermentation types were classified into ethanol-type, mixed acid-type, propionic acid-type, and butyric acid-type fermentations, which occurred at pH 4.0 ~ 4.5, 4.5 ~ 5.0, 5.0 ~ 5.5, and 5.5 ~ 6.5, respectively. [ABSTRACT FROM AUTHOR]

Background: Acute toxicity means the ability of a substance to cause adverse effects within a short period following dosing or exposure, which is usually the first step in the toxicological investigations of unknown substances. The median lethal dose, LD, is frequently used as a general indicator of a substance's acute toxicity, and there is a high demand on developing non-animal-based prediction of LD. Unfortunately, it is difficult to accurately predict compound LD using a single QSAR model, because the acute toxicity may involve complex mechanisms and multiple biochemical processes. Results: In this study, we reported the use of local lazy learning (LLL) methods, which could capture subtle local structure-toxicity relationships around each query compound, to develop LD prediction models: (a) local lazy regression (LLR): a linear regression model built using k neighbors; (b) SA: the arithmetical mean of the activities of k nearest neighbors; (c) SR: the weighted mean of the activities of k nearest neighbors; (d) GP: the projection point of the compound on the line defined by its two nearest neighbors. We defined the applicability domain (AD) to decide to what an extent and under what circumstances the prediction is reliable. In the end, we developed a consensus model based on the predicted values of individual LLL models, yielding correlation coefficients R of 0.712 on a test set containing 2,896 compounds. Conclusion: Encouraged by the promising results, we expect that our consensus LLL model of LD would become a useful tool for predicting acute toxicity. All models developed in this study are available via . [ABSTRACT FROM AUTHOR]

New Delhi metallo-β-lactamase-1 (NDM-1) has emerged as a major global threat to human health for its rapid rate of dissemination and ability to make pathogenic microbes resistant to almost all known β-lactam antibiotics. In addition, effective NDM-1 inhibitors have not been identified to date. In spite of the plethora of structural and kinetic data available, the accurate molecular characteristics of and details on the enzymatic reaction of NDM-1 hydrolyzing β-lactam antibiotics remain incompletely understood. In this study, a combined computational approach including molecular docking, molecular dynamics simulations and quantum mechanics/molecular mechanics calculations was performed to characterize the catalytic mechanism of meropenem catalyzed by NDM-1. The quantum mechanics/molecular mechanics results indicate that the ionized D124 is beneficial to the cleavage of the C-N bond within the β-lactam ring. Meanwhile, it is energetically favorable to form an intermediate if no water molecule coordinates to Zn2. Moreover, according to the molecular dynamics results, the conserved residue K211 plays a pivotal role in substrate binding and catalysis, which is quite consistent with previous mutagenesis data. Our study provides detailed insights into the catalytic mechanism of NDM-1 hydrolyzing meropenem β-lactam antibiotics and offers clues for the discovery of new antibiotics against NDM-1 positive strains in clinical studies. [ABSTRACT FROM AUTHOR]

The concept of drug-likeness has been widely applied in combinatorial chemistry as an approach to reduce attrition in drug discovery and development. Meanwhile, bicyclic privileged structures with versatile binding properties have emerged as ideal source of core scaffolds for the design and synthesis of combinatorial libraries. For the purpose of better assisting the design of bicyclic privileged structure-based combinatorial libraries, we conducted an integrated drug-likeness study on compounds of these scaffolds. Distributions of physicochemical properties (PCPs) were analyzed and in silico prediction models were built. Our results showed that there exist much difference between the drug-like ranges (DLRs) of bicyclic privileged structures and that of others, which have significant impact on compound selection. The DLRs for bicyclic privileged structures were defined as 260 ≤ MW ≤ 524; 0.9 ≤ ALogP ≤ 5.4; 2 ≤ Hacc ≤ 8; Hdon ≤ 3; 21.0 ≤ PSA ≤ 128.6; 6.3 ≤ FPSA ≤ 34.2; 1 ≤ RotB ≤ 10; 2 ≤ Nr ≤ 5; 1 ≤ Nc ≤ 7; SA ≤ 4. Two accurate and easy to understand in silico prediction models, Caco-2 permeability model and metabolic stability classification model, had been built to guide drug candidate optimization. In these models, hydrogen-bond donor and rotatable bond showed major impact on the permeability of compounds, while lipophilicity, flexibility, degree of branching and the existence of some functional groups determined the fate of a drug in metabolic process. Suggestions on structural modification toward higher permeability and metabolic stability were given according to the in silico models. [ABSTRACT FROM AUTHOR]

Skin sensitization is an important toxic endpoint in the risk assessment of chemicals. In this paper, structure-activity relationships analysis was performed on the skin sensitization potential of 357 compounds with local lymph node assay data. Structural fragments were extracted by GASTON (GrAph/Sequence/Tree extractiON) from the training set. Eight fragments with accuracy significantly higher than 0.73 ( p < 0.1) were retained to make up an indicator descriptor fragment. The fragment descriptor and eight other physicochemical descriptors closely related to the endpoint were calculated to construct the recursive partitioning tree (RP tree) for classification. The balanced accuracy of the training set, test set I, and test set II in the leave-one-out model were 0.846, 0.800, and 0.809, respectively. The results highlight that fragment-based RP tree is a preferable method for identifying skin sensitizers. Moreover, the selected fragments provide useful structural information for exploring sensitization mechanisms, and RP tree creates a graphic tree to identify the most important properties associated with skin sensitization. They can provide some guidance for designing of drugs with lower sensitization level. [ABSTRACT FROM AUTHOR]

The recent wide spreading of the H5N1 avian influenza virus (AIV) in Asia, Europe and Africa and its ability to cause fatal infections in human has raised serious concerns about a pending global flu pandemic. Neuraminidase (NA) inhibitors are currently the only option for treatment or prophylaxis in humans infected with this strain. However, drugs currently on the market often meet with rapidly emerging resistant mutants and only have limited application as inadequate supply of synthetic material. To dig out helpful information for designing potent inhibitors with novel structures against the NA, we used automated docking, CoMFA, CoMSIA, and HQSAR methods to investigate the quantitative structure–activity relationship for 126 NA inhibitors (NIs) with great structural diversities and wide range of bioactivities against influenza A virus. Based on the binding conformations discovered via molecular docking into the crystal structure of NA, CoMFA and CoMSIA models were successfully built with the cross-validated q 2 of 0.813 and 0.771, respectively. HQSAR was also carried out as a complementary study in that HQSAR technique does not require 3D information of these compounds and could provide a detailed molecular fragment contribution to the inhibitory activity. These models also show clearly how steric, electrostatic, hydrophobicity, and individual fragments affect the potency of NA inhibitors. In addition, CoMFA and CoMSIA field distributions are found to be in well agreement with the structural characteristics of the corresponding binding sites. Therefore, the final 3D-QSAR models and the information of the inhibitor–enzyme interaction should be useful in developing novel potent NA inhibitors. [ABSTRACT FROM AUTHOR]

A combinatorial pharmacophore (CP) model for Multidrug and toxin extrusion 1 (MATE1/SLC47A1) inhibitors was developed based on a data set including 881 compounds. The CP model comprises four individual pharmacophore hypotheses, HHR1, DRR, HHR2 and AAAP, which can successfully identify the MATE1 inhibitors with an overall accuracy around 75%. The model emphasizes the importance of aromatic ring and hydrophobicity as two important structural determinants for MATE1 inhibition. Compared with the pharmacophore model of Organic Cation Transporter 2 (OCT2/ SLC22A2), a functional related transporter of MATE1, the hypotheses of AAAP and PRR5 are suggested to be responsible for their ligand selectivity, while HHR a common recognition pattern for their dual inhibition. A series of analysis including molecular sizes of inhibitors matching different hypotheses, matching of representative MATE1 inhibitors and molecular docking indicated that the small inhibitors matching HHR1 and DRR involve in competitive inhibition, while the relatively large inhibitors matching AAAP are responsible for the noncompetitive inhibition by locking the conformation changing of MATE1. In light of the results, a hypothetical model for inhibiting transporting mediated by MATE1 was proposed. [ABSTRACT FROM AUTHOR]

Structure-based lead optimization is an open challenge in drug discovery, which is still largely driven by hypotheses and depends on the experience of medicinal chemists. Here we propose a pairwise binding comparison network (PBCNet) based on a physics-informed graph attention mechanism, specifically tailored for ranking the relative binding affinity among congeneric ligands. Benchmarking on two held-out sets (provided by Schrödinger and Merck) containing over 460 ligands and 16 targets, PBCNet demonstrated substantial advantages in terms of both prediction accuracy and computational efficiency. Equipped with a fine-tuning operation, the performance of PBCNet reaches that of Schrödinger's FEP+, which is much more computationally intensive and requires substantial expert intervention. A further simulation-based experiment showed that active learning-optimized PBCNet may accelerate lead optimization campaigns by 473%. Finally, for the convenience of users, a web service for PBCNet is established to facilitate complex relative binding affinity prediction through an easy-to-operate graphical interface., (© 2023. The Author(s).)