Search Results (36)

Background/Objectives: Endocrine resistance due to activating mutations in the estrogen receptor alpha (ERα), especially Y537S, is still a big challenge in the treatment of hormone receptor-positive breast cancer, and there is a need to develop new small-molecule inhibitors that can overcome ligand-independent receptor activation. We used an integrated machine learning and molecular dynamics (MD)-based virtual screening (VS) approach to identify potential computationally prioritized hits of ERα Y537S in this study. Methods: The dataset (~3353 compounds) was characterized using molecular fingerprints and graphically represented by PCA and t-SNE to show structurally different clusters linked to potency (pIC50). To guarantee robust downstream model training, the diagnosis of structural and influential outliers was completed through the application of rigorous data quality control methods, including the Williams plot, Mahalanobis distance, and Cook’s distance diagnostics. This was followed by multiple benchmarks of machine learning classifiers (random forest, SVM, KNN, and CNN) for activity classification, yielding the highest AUC of 0.95 for the random forest classifier. Interpretable structure–activity insights were gained by the identification of key substructures of the fingerprint by SHAP and feature importance analysis that influence the predicted potency. Three lead candidates (Hit-1, Hit-2, and Hit-3) were screened and tested with a 200 ns all-atom MD simulation in comparison with a reference control to evaluate the binding stability. Comprehensive trajectory analysis, such as PCA, FEL, RDF, salt bridges, and DCCM, was completed following the structure-based identification of the most stable ligands. Results: Three candidates were identified as Hit-1, Hit-2, and Hit-3. Hit-3 was found to be bound in the most stable binding pose and had the most similar conformational behavior to the control, whereas Hit-2 showed transient pose instability associated with increased anti-correlated domain-level motions and an alternate high-salt-bridge conformational state. As no computational metric can be an absolute measure of experimental affinity, Hit-1 has the most favorable calculated end-point binding-energy estimate, while Hit-3 showed conformational stability in MD simulations. Conclusions: The computationally prioritized candidates can be subjected to further experimental testing, as the conducted docking, MD simulations, and end-point free energy analyses cannot establish overall compound binding, cell-based activity, pharmacological mechanism, or therapeutic efficacy. Full article

Extended object tracking (EOT) has emerged as a critical challenge in modern surveillance and autonomous systems, where high-resolution sensors yield multiple measurements per target in a single scan. This task necessitates the simultaneous estimation of an object’s spatial extent and kinematic states by systematically assimilating distributed detections. In practical scenarios, however, sensor visibility constraints often restrict measurements to partial contours rather than the full object geometry, posing a significant hurdle for accurate state estimation. This paper proposes a robust EOT algorithm that addresses these challenges by extracting elliptical features—including size, orientation, and noise characteristics—in an affine-invariant manner from incomplete measurement sets. By formulating the ellipse-fitting problem through an algebraic solution, the proposed method effectively reconstructs the target’s morphology even under severe self-occlusion. The tracking framework is established as a Bayesian inference problem, utilizing a standard Kalman filter as a recursive estimator to ensure both computational efficiency and numerical robustness. In Monte Carlo simulations in which measurements are restricted to the sensor-facing contour, the proposed method reduces the average extent estimation error, quantified by the Gaussian Wasserstein distance, to 9.15 m, compared with 64.84–70.62 m for three state-of-the-art baselines, while retaining real-time feasibility with an average processing time of 0.94 ms per scan. Experiments on real-world LiDAR vehicle-tracking data further confirm consistent and accurate extent estimation under partial observability and complex maneuvers. Full article

Background: Intervertebral disc degeneration (IVDD) is driven by the interplay between inflammatory signaling, extracellular matrix (ECM) degradation, and impaired cellular adaptation. Although several nutraceutical compounds have been reported to exert protective effects in IVDD-related models, their multitarget mechanisms within integrated molecular networks remain incompletely characterized. Methods: An in silico framework integrating molecular docking with network-based analyses was employed to evaluate resveratrol, quercetin, melatonin, curcumin, and baicalein against a predefined panel of IVDD-associated targets, within an exploratory in silico framework. Binding affinities and interaction profiles were assessed using molecular docking, followed by protein–protein interaction (PPI) network construction, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, and hub gene identification. Results: Docking analyses revealed binding energies ranging from −4.59 to −13.25 kcal/mol, with curcumin and quercetin showing plausible docking poses across a subset of selected targets under the applied protocol. Network analysis showed a highly interconnected structure centered on key inflammatory regulators, including NFKB1, IL6, TNF, IL1B, STAT3, and NLRP3, together with ECM-associated components such as ACAN, COL2A1, SOX9, MMP13, and ADAMTS5. Enrichment analyses further suggested significant associations with inflammatory signaling pathways, cytokine regulation, and ECM organization. Conclusions: These findings are compatible with a distributed, multitarget interaction pattern of nutraceutical compounds within IVDD-associated molecular networks. By integrating molecular docking with network-based analyses, this study offers a system-level framework for interpreting previously reported effects within a disease-specific context. Docking-derived interaction patterns should be interpreted as qualitative and exploratory observations, as docking scores represent model-dependent estimates and do not establish comparable pharmacological effects across heterogeneous targets. The results should be considered hypothesis-generating and require experimental validation. Full article

Accurate alignment of real-world object poses with their virtual counterparts using sensors, e.g. cameras, is essential for consistent interaction in mixed-reality systems. However, objects can undergo abrupt, untracked movements during periods when a tracking system is inactive, e.g., overnight, causing stored pose records to become inconsistent with the real scene and breaking user interaction in the virtual environment. Off-the-shelf 3D reconstruction networks such as MASt3R (Matching and Stereo 3D Reconstruction) method provide metrically scaled 3D point maps and pixel correspondences, but they are trained on static scenes and therefore fail to produce reliable object correspondences when the object has moved. We propose a robust pipeline that combines MASt3R’s metrically scaled 3D outputs with a background-based alignment strategy to recover and apply the true pose change of moved objects. Our method first segments foreground and background and extracts 3D background point sets for a reference day and a current day. An affine transformation between these background point sets is estimated via a standard registration technique and used to express the current-day object 3D coordinates in the reference coordinate frame. Within that unified frame we compute the object pose change and apply the resulting transform to the virtual object, restoring real–virtual consistency. Experiments on real scenes demonstrate that the proposed approach reliably corrects pose misalignments introduced during inactive periods and substantially improves over applying MASt3R alone, thereby enabling restored and consistent user interaction in the virtual environment. Full article

Gaze estimation plays a crucial role in human-computer interaction and behavior analysis. However, in dynamic scenes, rigid head movements and rapid gaze shifts pose significant challenges to accurate gaze prediction. Most existing methods either process single-frame images independently or rely on long video sequences, making it difficult to simultaneously achieve strong performance and high computational efficiency. To address this issue, we propose DGAGaze, a gaze estimation framework based on a difference-driven spatiotemporal attention mechanism. This framework uses a geometry-aware temporal alignment module to mitigate interference from rigid head movements, compensating for them through pose estimation and affine feature warping, thereby achieving explicit decoupling between global head motion and local eye motion. Based on the aligned features, inter-frame differences are used to adjust spatial and channel attention weights, enhancing motion-sensitive representations without introducing an additional temporal modeling layer. Extensive experiments on the EyeDiap and Gaze360 datasets demonstrate the effectiveness of the proposed approach. DGAGaze achieves improved gaze estimation accuracy while maintaining a lightweight architecture based on a ResNet-18 backbone, outperforming existing state-of-the-art methods. Full article

Visual localization plays a critical role for mobile robots to estimate their position and orientation in GPS-denied environments. However, its efficiency, robustness, and generalization are fundamentally undermined by severe viewpoint changes and dramatic appearance variations, which present persistent challenges for image-based feature representation and pose estimation under real-world conditions. Recently, map-free visual relocalization (MFVR) has emerged as a promising paradigm for lightweight deployment and privacy isolation on edge devices, while how to learn compact and invariant image tokens without relying on structural 3D maps still remains a core problem, particularly in highly dynamic or long-term scenarios. In this paper, we propose the Debiased Multiplex Tokenizer as a novel method (termed as DMT-Loc) for efficient and versatile MFVR to address these issues. Specifically, DMT-Loc is built upon a pretrained vision Mamba encoder and integrates three key modules for relative pose regression: First, Multiplex Interactive Tokenization yields robust image tokens with non-local affinities and cross-domain descriptions. Second, Debiased Anchor Registration facilitates anchor token matching through proximity graph retrieval and autoregressive pointer attribution. Third, Geometry-Informed Pose Regression empowers multi-layer perceptrons with a symmetric swap gating mechanism operating inside each decoupled regression head to support accurate and flexible pose prediction in both pair-wise and multi-view modes. Extensive evaluations across seven public datasets demonstrate that DMT-Loc substantially outperforms existing baselines and ablation variants in diverse indoor and outdoor environments. Full article

The aim of this study was to investigate the molecular interactions of heparin, diclofenac, and their supramolecular complexes with cyclooxygenase enzymes (COX-1 and COX-2) using computational docking techniques. Diclofenac is a widely used nonsteroidal anti-inflammatory drug (NSAID) that inhibits COX isoforms, whereas heparin is a polyanionic glycosaminoglycan with established anticoagulant and emerging anti-inflammatory properties. Supramolecular association between these agents may modulate their physicochemical behavior and target engagement. Molecular modeling, dual-drug docking, and molecular dynamics (MD) simulations were employed to characterize the interactions of heparin, diclofenac, and pre-formed heparin–diclofenac complexes with COX-1 and COX-2. Geometry optimization and lipophilicity (logP) estimates were obtained using HyperChem, while protein–ligand docking was performed in HEX using crystallographic COX structures from the Protein Data Bank. Docking poses were analyzed in Chimera, and selected complexes were refined through short MD simulations. Pre-formed heparin–diclofenac assemblies exhibited markedly enhanced docking scores toward both COX isoforms compared with single ligands. Binding orientation strongly influenced affinity: for COX-1, the heparin–diclofenac configuration yielded the most favorable interaction, whereas for COX-2 the diclofenac–heparin configuration was preferred. Both assemblies adopted binding modes distinct from free diclofenac, suggesting cooperative electrostatic and hydrophobic contacts at the enzyme surface. Supramolecular complexation also altered calculated logP values relative to the individual compounds. MD simulations supported the relative stability of the top-ranked complex–COX assemblies. These findings indicate that heparin–diclofenac assemblies may enhance and reorganize predicted COX interactions in a configuration-dependent manner and illustrate the utility of dual-drug docking for modeling potential synergistic effects. Such insights may inform the design of localized or topical formulations, potentially incorporating non-anticoagulant heparin derivatives, to achieve effective COX inhibition with reduced systemic exposure. However, the results rely on simplified heparin fragments, legacy docking tools, and short MD simulations, and should therefore be interpreted qualitatively. Experimental studies will be essential to confirm whether such supramolecular assemblies form under physiological conditions and whether they influence COX inhibition in vivo. Full article

Introduction: Acinetobacter baumannii is a multidrug-resistant pathogen from the ESKAPE group, associated with nosocomial infections. Its resistance to multiple antibiotics poses a global threat. The QACE protein, an efflux pump, has been identified as a key resistance mechanism, making it a promising target for the development of new antibacterial agents. Objective: Our aim is to identify low-molecular-weight compounds derived from natural products with potential inhibitory activity against the QACE protein, using virtual screening and molecular docking studies. Materials and Methods: The three-dimensional structure of QACE was retrieved from AlphaFold, followed by energy minimization and assignment of Kollman-type charges. Ligand screening was performed using the BioMX database through structural similarity analysis (Tanimoto coefficient ≥ 0.85), using ciprofloxacin as the reference compound. Selected molecules were evaluated using SwissADME to predict their pharmacokinetic properties, and three candidates with favorable profiles were chosen. Molecular docking studies were then performed using AutoDock 4 to estimate binding affinities. Results: Voacangine was the compound with the highest structural similarity, strongest binding affinity to QACE (ΔG = −6.2 kcal/mol; ki = 28.51 μM), and stable molecular interactions including hydrogen bonds and π-stacking. It showed favorable tissue distribution, low potential for CYP3A4 inhibition, and minimal predicted cardiotoxicity (hERG channel blockade). Conclusion: Voacangine emerges as a promising candidate for inhibiting the QACE efflux pump in Acinetobacter baumannii. This study highlights the value of computer-aided drug design as an effective strategy in the search for new treatments against multidrug-resistant bacteria. Full article

Three-dimensional fragment assembly technology has significant application value in fields such as cultural relic restoration, medical image analysis, and industrial quality inspection. To address the common challenges of limited feature representation ability and insufficient assembling accuracy in existing methods, this paper proposes a geometry-aware hierarchical fragment assembly framework (GeoAssemble). The core contributions of our work are threefold: first, the framework utilizes DGCNN to extract local geometric features while integrating centroid relative positions to construct a multi-dimensional feature representation, thereby enhancing the identification quality of fracture points; secondly, it designs a two-stage matching strategy that combines global shape similarity coarse matching with local geometric affinity fine matching to effectively reduce matching ambiguity; finally, we propose an auxiliary transformation estimation mechanism based on the geometric center of fracture point clouds to robustly initialize pose parameters, thereby improving both alignment accuracy and convergence stability. Experiments conducted on both synthetic and real-world fragment datasets demonstrate that this method significantly outperforms baseline methods in matching accuracy and exhibits higher robustness in multi-fragment scenarios. Full article

Background: Hydroxycarboxylic acid receptor 3 (HCAR3) is a receptor that is mainly expressed in human adipose tissue. It can inhibit lipolysis through the inhibition of adenylyl cyclase; thus, it is closely related to the regulation of lipids in the human body. This makes HCAR3 a compelling target for developing drugs against dyslipidemia. Notably, the reported active compounds for HCAR3 are all carboxylic acids. This observation is in line with the fact that ARG111 has been reported as the key residue to anchor the active compound in a closely related homologous protein—HCAR2. Methods: In this study, we aim to discover new chemicals, through virtual screening, that may bind with HCAR3. As there are several choices for the receptor conformation, cross-docking was conducted and the root-mean-square deviation of the docking pose from the conformation of the crystal ligand was employed to determine the best receptor conformation for screening. Ligands from the ZINC20 database were screened through molecular docking, and 30 candidates were subjected to 100 ns MD simulations. Six stable complexes were further assessed by umbrella sampling to estimate binding affinity. Results: The homology model (HCAR3_homology) was selected as the receptor. Following the protocol determined by the retrospective docking process, prospective docking was conducted to screen the ligands from the ZINC20 database. Subsequently, the top 30 compounds with a good docking score and a good interaction with ARG111 were subjected to 100 ns molecular dynamics (MD) simulations, and their binding stability was analyzed based on the resulting trajectories. Finally, six compounds were chosen for binding free energy calculation using umbrella sampling; all showed negative binding affinities. Conclusions: All six compounds selected for umbrella sampling showed negative binding affinities, suggesting their potential as novel HCAR3 ligands for the development of drugs against dyslipidemia. Full article

The SARS-CoV-2 spike receptor-binding motif is crucial for viral entry via interaction with the human ACE2 receptor. Mutations N501Y and E484K, found in several variants of concern, impact viral transmissibility and immune escape, but experimental data on their binding effects remain inconsistent. Using isothermal titration calorimetry (ITC) and molecular dynamics (MD) simulations, we analyzed the thermodynamic and structural effects of these mutations. ITC confirmed that N501Y increases ACE2 affinity by 2.2-fold, while E484K enhances binding by 5.8-fold. The Beta/Gamma variant (carrying both mutations) showed the strongest affinity, with a 15-fold increase. E484K was enthalpy-driven, while N501Y introduced entropy-driven effects, suggesting hydrophobic interactions and conformational changes. MD simulations revealed distinct binding poses, with Beta/Gamma peptides interacting with a secondary ACE2 site. A strong correlation was found between entropy contributions and hydrophobic contacts. Additionally, a convolutional neural network was used to estimate the free binding energy of these complexes. Our findings confirm that N501Y and E484K enhance ACE2 binding, with the greatest effect when combined, providing insights into SARS-CoV-2 variant evolution and potential therapeutic strategies. Full article

Background/Objectives: Influenza viruses are highly transmissible and mutable, posing a significant burden on public health. This study aimed to design a recombinant multi-epitope vaccine with broad protective potential. Methods: Immunoinformatic approaches were employed to predict epitopes from over 30,000 protein sequences retrieved from protein databases. Epitopes were filtered using four key indicators: antigenicity, allergenicity, toxicity, and conservancy. Population coverage analysis was conducted to estimate the proportion of the global population that could potentially benefit from the vaccine. Secondary and tertiary structures of the recombinant vaccine were predicted using the PSIPRED server and AlphaFold2. The vaccine efficacy was validated through an immune simulation, molecular docking, and molecular dynamics simulation. Results: A recombinant multi-epitope vaccine demonstrating strong antigenicity, no allergenicity or toxicity, and high conservation across different subtypes was successfully constructed. Population coverage analysis indicated that the vaccine could elicit an immune response in 90.14% of the global population. Both the secondary and tertiary structures of the vaccine were accurately predicted. Molecular dynamics simulations further validated the structural stability and interactions of the vaccine components with TRL4. Molecular docking confirmed the robust binding affinity of T-cell epitopes to MHC molecules. Simulated immunity studies showed that the vaccine induced the proliferation of memory B cells and T cells, enabling rapid antibody production during viral challenges. Conclusions: This study provides a promising basis for the development of a broadly protective influenza vaccine, leveraging cutting-edge immunoinformatics and molecular dynamics simulations to address the global challenge posed by influenza virus variability. Full article

Environmental pollution from organic dyes such as malachite green and rhodamine B poses significant threats to ecosystems and human health due to their toxic properties. The rapid detection of these contaminants with high sensitivity and selectivity is crucial and can be effectively achieved using nonlinear optical methods. In this study, we combine the Simplified Bond Hyperpolarizability Model (SBHM) and molecular docking (MD) simulations to investigate the Second-Harmonic Generation (SHG) intensity of organic dyes on a silicon (Si(001)) substrate for nanoscale pollutant detection. Our simulations show good agreement with rotational anisotropy (RA) SHG intensity experimental data across all polarization angles, with a total error estimate of 3%. We find for the first time that the SBHM not only identifies the different organic pollutant dyes on the surface, as in conventional SHG detection, but can also determine their relative orientation and different concentrations on the surface. Meanwhile, MD simulations reveal that rhodamine B shows a strong adsorption affinity of $-10.4\mathrm{kcal}/\mathrm{mol}$ to a single-layer graphene oxide (GO) substrate, primarily through $\pi$-$\pi$ stacking interactions (36 instances) and by adopting a perpendicular molecular orientation. These characteristics significantly enhance SHG sensitivity. A nonlinear susceptibility analysis reveals good agreement between the SBHM and group theory. The susceptibility tensors confirm that the dominant contributions to the SHG signal arise from both the molecular structure and the surface interactions. This underscores the potential of GO-coated silicon substrates for detecting trace levels of organic pollutants with interaction distances ranging from $3.75\AA$ to $5.81\AA$. This approach offers valuable applications in environmental monitoring, combining the sensitivity of SHG with the adsorption properties of GO for nanoscale detection. Full article

At present, Unmanned Aerial Vehicles (UAVs) combined with deep learning have become an important means of transmission line inspection; however, the current approach has the problems of high demand for manual operation, low inspection efficiency, inspection results that do not reflect the distribution of defects on transmission towers, and the need for a large number of manually annotated captured images. In order to achieve the UAV understanding the structure of transmission towers and identifying the defects in the parts of transmission towers, a three-granularity pose estimation framework for multi-type high-voltage transmission towers using Part Affinity Fields (PAFs) is presented here. The framework classifies the structural critical points of high-voltage transmission towers and uses PAFs to provide a basis for the connection between the critical points to achieve the pose estimation for multi-type towers. On the other hand, a three-fine-grained prediction incorporating an intermediate supervisory mechanism is designed so as to overcome the problem of dense and overlapping keypoints of transmission towers. The dataset used in this study consists of real image data of high-voltage transmission towers and complementary images of virtual scenes created through the fourth-generation Unreal Engine (UE4). In various types of electrical tower detection, the average keypoint identification AF of the proposed model exceeds 96% and the average skeleton connection AF exceeds 93% at all granularities, which demonstrates good results on the test set and shows some degree of generalization to electricity towers not included in the dataset. Full article

Precise binding free-energy predictions for ligands targeting metalloproteins, especially zinc-containing histone deacetylase (HDAC) enzymes, require specialized computational approaches due to the unique interactions at metal-binding sites. This study evaluates a docking algorithm optimized for zinc coordination to determine whether it could accurately differentiate between protonated and deprotonated states of hydroxamic acid ligands, a key functional group in HDAC inhibitors (HDACi). By systematically analyzing both protonation states, we sought to identify which state produces docking poses and binding energy estimates most closely aligned with experimental values. The docking algorithm was applied across HDAC 2, 4, and 8, comparing protonated and deprotonated ligand correlations to experimental data. The results demonstrate that the deprotonated state consistently yielded stronger correlations with experimental data, with R² values for deprotonated ligands outperforming protonated counterparts in all HDAC targets (average R² = 0.80 compared to the protonated form where R² = 0.67). These findings emphasize the significance of proper ligand protonation in molecular docking studies of zinc-binding enzymes, particularly HDACs, and suggest that deprotonation enhances predictive accuracy. The study’s methodology provides a robust foundation for improved virtual screening protocols to evaluate large ligand libraries efficiently. This approach supports the streamlined discovery of high-affinity, zinc-binding HDACi, advancing therapeutic exploration of metalloprotein targets. A comprehensive, step-by-step tutorial is provided to facilitate a thorough understanding of the methodology and enable reproducibility of the results. Full article