Search Results (3,321)

Drug discovery is still an expensive and time-consuming process where finding the right drug associations is important for therapeutic development. In this paper, a new system is proposed for drug design called PharmaGraphFormer (PGF). It consists of five stages: (i) Data acquisition and preprocessing (DAP), (ii) Feature extraction and feature fusion (FEF), (iii) Molecular representation (MR), (iv) Multi-task prediction, and (v) Explainable artificial intelligence (XAI). This study employs a hybrid graph neural network (GNN)-transformer architecture that combines structural and sequence-based representations. Through DAP, several processes are executed, including the imputation or removal of missing values, outlier rejection, and class balancing. Next, through FEF1, features are extracted to represent the input data efficiently. Initially, compound-protein features are generated to document the interactions and relationships between chemical compounds and their corresponding target proteins. Secondly, drug characterizations are computed to encapsulate the physical, chemical, and structural attributes of each drug. After that, MR is performed using a graph-based molecule representation. Then, a novel model integrating GNNs and graph transformers, termed GNN-T, is proposed. Initially, GNNs represent the most promising deep learning models adept at processing non-Euclidean data. The Graph Transformer layer enhances atom representations by consolidating the representations of adjacent atoms through an attention mechanism. Finally, XAI is applied to explain the internal mechanisms of AI systems, rendering them comprehensible and interpretable. Across five independent runs, the proposed model achieved an accuracy of $0.963\pm 0.002$, a precision of $0.971\pm 0.002$, a recall of $0.958\pm 0.003$, an F1-score of $0.964\pm 0.002$, and a ROC-AUC of $0.993\pm 0.001$. These results demonstrate an outstanding performance when compared with all other models and emphasize that the proposed model is reliable in solving the problems of prioritizing compounds in line with the latest developments in AI-powered virtual screening and drug–target interaction modeling. Full article

Under ground-based observation geometry, infrared radiant-intensity sequences of space objects are jointly influenced by object micromotion, thermal radiation, time-varying viewing conditions, and atmospheric propagation. Existing simulation studies often prescribe the line of sight or simplify the coupling between viewing geometry and atmospheric attenuation, which limits long-duration ground-based sequence analysis. This study develops a physics-informed framework for generating atmosphere-attenuated infrared radiant-intensity sequences of space objects undergoing precession or tumbling. The framework reconstructs observation geometry from azimuth–elevation–range trajectories, updates facet normals through a unified micromotion attitude model, computes visible projected area and transient facet temperature, and incorporates MODTRAN-derived elevation-dependent atmospheric transmittance. Using this framework, we construct IRPeriodic, an eight-class simulated dataset for long-duration univariate time-series classification. We further propose LPD-Net, which integrates large-kernel residual feature extraction, prototype-guided dynamic temporal alignment, and differential periodic representation to capture long-range waveform morphology, sample-dependent temporal correspondence, and segment-level local variation. On IRPeriodic, LPD-Net achieves an accuracy of 0.8618 ± 0.0057, a macro-F1 of 0.8615 ± 0.0061, and a Matthews correlation coefficient of 0.8426 ± 0.0065, outperforming the evaluated neural-network and ROCKET-type baselines. Ablation and synthetic-noise sensitivity analyses indicate that the performance gain is mainly associated with long-context feature extraction, with additional improvements from dynamic alignment and differential periodic statistics. Auxiliary experiments on selected public UCR datasets suggest that the representation is also competitive for univariate time-series classification. These results demonstrate the effectiveness of LPD-Net on the proposed physics-informed benchmark for long-duration ground-based infrared radiant-intensity sequence classification. Full article

To address output-side high-frequency harmonics in a three-phase-to-single-phase matrix converter (3-1MC) with an inductive compensation unit and topological port asymmetry, two PWM switching-sequence optimization methods are proposed. Without power decoupling, the pulsating power associated with the single-phase output is coupled to the input side through the bidirectional switching network because the converter has no large energy-storage DC link. Under conventional modulation, the state sequence can produce large output-voltage steps and nonuniform commutation paths, thereby increasing switching-frequency harmonic components. The first proposed method avoids direct commutation of the line voltage with the largest instantaneous magnitude to the zero state by inserting line-voltage segments with smaller instantaneous magnitudes. The second method rearranges the switching-state sequence without changing the effective-vector durations, thereby reducing the number of switching transitions within each switching cycle. Compared with the conventional modulation method, Method 1 reduces the output-voltage THD from 35.4% to 31.4%, corresponding to a relative reduction of 11.3%. Method 2 reduces the THD from 35.4% to 27.5%, corresponding to a relative reduction of 22.3%. Full article

The slow evolution and inconspicuous nature of buffer-layer ablation in high-voltage cables pose a significant challenge for early fault diagnosis. To tackle this issue, we propose a hybrid diagnostic approach that integrates frequency-domain impedance measurement with a convolutional neural network (CNN). A cable simulation model is first established using transmission-line theory and a distributed-parameter framework. We examine the impedance and phase responses at the cable’s sending end, revealing a consistent decreasing trend with rising frequency alongside periodic resonant peaks. The simulator generates a diverse set of spectral signatures corresponding to various cable health states. The CNN then extracts discriminative features from these waveforms, and a probabilistic clustering preprocessing step further refines the input data. Experimental results on a test set of 78 samples—comprising 52 experimentally measured normal spectra and 26 experimentally calibrated simulated spectra for mild and severe ablation—demonstrate a classification accuracy of 0.95, confirming that the proposed methodology enables reliable, non-intrusive detection of buffer-layer ablation without cable disassembly. Full article

This paper proposes Non-Uniform Adaptive Acquisition (NUAA), a signal-feature-matched non-uniform adaptive photonic sampling framework that recovers broadband radio-frequency (RF) waveforms from highly sparse programmable non-uniform photonic sampling points. A 200 MHz mode-locked laser together with five electrical optical delay lines (EDLs; motor-actuated optical delay units) arranges the non-uniform sampling instants. Benefiting from the joint design of the photodetector/track-and-hold amplifier (PD/THA) response model and programmable non-uniform optical pulse spacing, a low-bandwidth PD infers neighboring pulse amplitudes from their deterministic superposition at the readout. In numerical simulations of this physical forward operator, that construction corresponds to a 1 THz equivalent sampling rate on the 1 ps EDL grid, while the electrical front end operates at a 1 GHz average sampling rate (cascaded PD–THA analog 3 dB bandwidth $\approx 0.676$ GHz). Under severe blocking interference and low signal-to-noise ratio (SNR), the numerical simulations show that the strongest broadband chirplet result uses a scene prior with support locking: with the NUAA–MU (Mamba–Unfolding) reconstructor at 0.1% multi-coset sparsity, all $N_{\mathrm{trial}}=50$ Monte Carlo trials succeed within 200 ms (Wilson 95% CI $[93,100]\%$; cumulative-best NMSE $-28.0$ dB), whereas the configuration without a scene prior is substantially weaker in the same window. A scene prior may come from known radar or communication waveform families, coarse occupancy reported by a companion sensor, or accumulation across related tasks. Full article

UAV-based insulator detection is challenging because insulators often occupy small regions of aerial images and appear against complex backgrounds, while subtle local features may be lost during feature extraction and down-sampling. Moreover, practical UAV and edge-device applications require efficient models with limited computational and memory demands. This study proposes an optimized lightweight YOLOv11n model for high-voltage transmission-line insulator detection. The architecture integrates C3k2MBNV2 to reduce model complexity, SCDown to preserve spatial information during down-sampling, and C3k2WTDC to enhance multi-frequency feature representation. A diverse dataset containing 5750 insulator images acquired under different environmental conditions, viewing angles, and backgrounds was used for evaluation. Experimental results show that the proposed model reduces the parameter count from 6.20 M to 3.26 M and computational complexity from 20.5 to 12.7 GFLOPs, corresponding to reductions of 47.4% and 38.0%, respectively. Meanwhile, precision increases from 91.3% to 93.8%, recall from 73.4% to 75.2%, mAP50 from 71.2% to 73.9%, and mAP50–95 from 65.6% to 67.3%. These results demonstrate an improved accuracy–efficiency trade-off, supporting real-time insulator detection in resource-constrained UAV and edge-device applications. Full article

Against the backdrop of the dual-carbon goals, a new power system is undergoing rapid construction, and distributed renewable energy is being connected at high density to regional ring networks. This study deeply explores the impact of photovoltaic (PV) power station connection positions on energy distribution and flow, as well as the voltage and energy loss of ring networks. Firstly, it takes the actual 220 kV/500 kV ring network in a certain city in Yunnan Province as the research object, and constructs a high-precision ETAP simulation model. It then systematically explores the mechanism of how PV power connection positions and grid connection penetration rates affect node voltage and line loss and derives the energy loss calculation formula. On this basis, two differentiated operation scenes corresponding to renewable energy output peaks and valleys are established. Through comparative analysis of multiple sets of simulation data, this study reveals the coupling laws among PV output fluctuation, bidirectional reverse power flow, system energy loss, and voltage over-limit. Finally, combined with the operational pain points of existing ring network and distribution network connection modes, this study proposes diversified loss reduction optimization strategies, including coordinated optimization of active and reactive power, and coordinated regulation of PV power and energy storage. The relevant research conclusions and optimization methods can improve the line loss analysis theory for ring networks integrated with PV power, and provide important engineering references for renewable energy planning and design, operation regulation, and loss management for similar mountain power grids. Full article

The deployment of deep-learning vulnerability detectors in production remains difficult. Models are large, false-positive rates are high, output is opaque, and a persistent gap separates benchmark performance from real-world utility. The objective of this work is to close part of that gap by combining a learned detector with interpretable rules so that accuracy, efficiency, and actionability are achieved together. A hybrid framework is therefore presented in which a lightweight edge-conditioned GNN of 71,810 parameters, named FastVulnGNN, trained in 96.2 s on a single CPU core, is paired with rule-based heuristic detection for six C/C++ vulnerability classes, namely buffer overflows, format-string defects, null-pointer dereferences, double-free errors, integer overflows, and race conditions. On the MegaVul dataset, an accuracy of 71.1%, an F1 score of 0.70, and an AUC-ROC of 0.77 are obtained by the GNN component. On a production codebase of 499 files and 312,758 lines of code, the full hybrid scan completes in 5.5 s, which corresponds to about 57,000 lines per second, without any GPU hardware. Per-file risk tiers and pattern-level explanations are produced, and these are suitable for continuous-integration use. The significance of this work lies in demonstrating that a deployable, explainable detector can be assembled from compact components, and an edge-type ablation study, a cross-dataset evaluation, and a per-vulnerability analysis are reported to characterize the approach. Full article

The anterior nucleus of the thalamus (ANT) is a key node of the hippocampal–diencephalic memory system and an established target for deep brain stimulation in drug-refractory epilepsy, yet its temporal connectivity remains incompletely characterized. This study delineates a dorsal posterolateral fiber bundle between the ANT and the amygdalo-uncal region, for which we propose the term anterior thalamo-temporal projection (ATTP). Fourteen hemispheres from seven human brains were examined using Klingler fiber dissection. The ATTP was identified in all 14 dissected hemispheres, following a dorsal posterolateral course through the pulvinar–lateral geniculate corridor and forming one macroscopic endpoint at the ANT, with no demonstrable ventral continuation. The reproducibility of a corresponding tractographic pattern was evaluated in 20 hemispheres from 10 Human Connectome Project participants using a repeated-run diffusion tractography protocol. An anatomically concordant tractographic pattern was reconstructed in 17 of 20 hemispheres (85%). The position of the ANT relative to the anterior commissure–posterior commissure line was also measured. The ANT position varied predominantly along the anteroposterior axis. To our knowledge, this study provides the first systematic microsurgical and tractographic delineation of the ATTP, establishing an anatomical framework for future investigations of medial temporal connectivity and patient-specific ANT neuromodulation. Full article

Background: Dry eye disease (DED) impairs quality of life. Artificial tears are the first-line treatment, but compliance is generally poor. Patients forget to instil drops or underestimate the importance of regular use. Mobile health (mHealth) applications improve adherence in other chronic conditions, but their role in DED remains unclear. This study evaluated the effectiveness of a smartphone reminder application in improving compliance to a four-times-daily artificial tear regimen and the associated symptom relief among patients with DED. Methods: A masked randomised crossover trial was conducted in 29 women with DED (mean ± SD age 21.2 ± 3.0 years). Participants completed two 14-day study phases: one with app-based reminders and one without, separated by a 7-day washout period. The primary outcome was mean daily drop frequency. Secondary outcomes were self-reported symptom frequency and severity assessed using the Symptom Assessment iN Dry Eye (SANDE). Results: The mean daily drop frequency was higher during the app phase compared with the non-app phase (F = 22.906, p < 0.001) but declined in both phases over time (F = 3.023, p < 0.001). Symptom frequency did not differ between phases but decreased over time (F = 1.993, p = 0.021). Symptom severity remained unchanged with no significant effects by phase or time. Baseline clinical measures did not predict drop use frequency or app-related improvement (p > 0.05). Conclusions: The reminder application increased short-term compliance to artificial tear use, but there was no corresponding reduction in symptoms or signs over two weeks’ usage. Compliance waned over time despite active reminders, suggesting that prompts alone are insufficient for sustaining behaviour change. Future digital interventions for DED should incorporate strategies to enhance motivation, support long-term engagement, and provide personalised education. Full article

Accurate identification of the initial traveling wavefront is essential for two-terminal fault location in modular multilevel converter-based high-voltage direct-current (MMC-HVDC) lines. This study develops a coordinated processing chain that combines a signed-pole modal transformation, offline whale optimization algorithm (WOA) calibration of variational mode decomposition (VMD), wavefront-sensitive intrinsic mode function (IMF) pairing, wavelet synchrosqueezing transform (WSST) diagnostics, and fractional-delay cross-correlation. A common VMD parameter pair is calibrated for both terminals. Candidate IMFs are then screened using wavefront-retention and frequency-consistency constraints. The final time difference is estimated from three fixed correlation windows after band-limited resampling and local parabolic peak refinement. This procedure estimates a sub-sample delay but does not increase the measurement bandwidth. The method is evaluated on a 200 km, 500 kV MMC-HVDC simulation model. The 10 kHz validation matrix contains 36 combinations of four fault types, three fault locations, and three fault resistances. The mean absolute error is 0.374 km, 34 of 36 errors are below 1 km, and the maximum error is 0.920% of the line length. A separate 10/50/100 kHz study evaluates the complete processing chain at all three rates. At the 100 km reference location, the complete method gives a mean absolute error of 0.148 km. The corresponding values are 0.213 km without modal transformation, 11.062 km with fixed VMD parameters, and 0.414 km when CWT replaces WSST. Direct line-mode and WTMM sample-level baselines each give a three-location mean absolute error of 3.433 km. The results support the coordinated design within the tested simulation envelope; noise, synchronization, wave-speed uncertainty, and near-terminal faults require further validation. A six-case robustness matrix with 540 noise realizations gives valid-location rates of 94.4%, 92.8%, and 89.4% at 40, 30, and 20 dB, respectively, with accepted-case median errors of 0.170, 0.181, and 0.270 km. Controlled synchronization, wave-speed, and combined-uncertainty tests further quantify the applicable error envelope. Full article

A series of pyruvamide 4-allylthiosemicarbazones HL^1–6 and their copper(II) chloride complexes 1–6 were synthesized. Their composition and structures were studied using elemental analysis, ¹H and ¹³C NMR spectroscopy, and FTIR spectroscopy, as well as single-crystal X-ray diffraction analysis for thiosemicarbazones HL², HL⁴, and HL⁶. The antiproliferative activity was evaluated against three human cancer cell lines. The copper(II) complexes exhibit higher anticancer activity than the corresponding thiosemicarbazones. Complexes 1 and 3 demonstrate the most promising anticancer properties and, in some cases, surpass the activity of doxorubicin, the reference drug. Their selectivity indices toward human leukemia THP-1 cells relative to the normal hTERT-RPE1 cell line are greater than 10. The copper(II) coordination compounds also exhibit moderate antibacterial and antifungal properties; however, this activity does not exceed that of the standard drugs. In contrast, thiosemicarbazones HL^1–6 show high antiradical activity, which surpasses that of both the corresponding copper(II) complexes and Trolox. All studied biological activities strongly depend on the nature of the substituent at the nitrogen atom of the pyruvamide moiety, highlighting the potential for further exploration of substituted pyruvamides with improved activity. Full article

Mixing a proppant-laden base fluid with high-pressure CO₂ can reduce its effective sand volume fraction, thereby impairing proppant placement and fracture conductivity. This study performs a numerical structural screening of the helical guide groove in a centrifugal–filtration sand concentrator intended for the pre-foaming base-fluid line upstream of the foam generator and operating against a 105 MPa outlet backpressure. The full-scale base geometry—an inner diameter of 500 mm, a concentration-zone length of 3500 mm, and 18 slotted-screen openings—was retained, while groove depth (10–40 mm), groove width (110–150 mm), and pitch (300–700 mm) were varied sequentially in 14 factor-level evaluations (12 unique geometries). Steady-state calculations were conducted in ANSYS Fluent 2022 R1 using an Eulerian–Eulerian two-fluid framework, the RNG k–ε turbulence closure, a constant apparent liquid viscosity of 0.040 Pa·s, and a compiled UDF-based phase-selective screen condition. CO₂ was not included as a separate phase. Within the investigated ranges, increasing groove depth and width increased the sand volume fraction at the concentrated-fluid outlet, whereas pitch produced a pronounced inverted-U-shaped response. The best tested geometry had a pitch of 500 mm, a width of 150 mm, and a depth of 40 mm, yielding an outlet sand volume fraction of 54.05%. This value is 3.80 percentage points higher than the lowest-performing tested configuration (50.25%), corresponding to a relative increase of 7.56%; it is also 14.05 percentage points above the inlet value of 40.00% and 4.05 percentage points above the engineering target of 50.00%. The results provide comparative numerical evidence for groove-parameter selection; independent experimental validation remains required before field application. Full article

A new mechanism for the formation of spectral–spatial distortions in cosmic microwave background radiation near primordial massive compact objects (such as primordial black holes) at redshifts from $z\sim 1000$ to ∼100 is proposed. After hydrogen recombination, the radiation from these objects leads to a significant increase in the population of hydrogen subordinate levels in their surrounding environment. Consequently, this allows for the observation of a Fraunhofer-like absorption spectrum in the cosmic microwave background. Such a distortion is formed due to the temperature difference between matter and relic radiation at the corresponding epoch. Ultimately, we should observe circular absorption or emission features around the objects with small angular sizes. These subordinate hydrogen lines currently lie in the radio wavelength range. Estimates indicate that the effect under consideration is accessible for the observations with planned large radiotelescopes. The possibility of the proposed mechanism lies in the ability of an object (e.g., accretion disk around black hole) to emit few-eV photons that populate the $n=2,3,4$ and higher levels of hydrogen, enabling the required excitation. Full article

Tropospheric evaporation ducts can trap radio waves within a refractive layer, which may guide signals above 1 GHz, enabling beyond-line-of-sight transmission. This makes duct-assisted propagation attractive for maritime communications. The marine environment is characterized by complex hydrometeorological variability and frequent extremes, particularly tropical cyclones, which can perturb duct properties and degrade link reliability. This study develops a multivariate cyclone-aware nonlinear regression framework (CNRF) to estimate contemporaneous evaporation duct strength (EDS) by integrating high-resolution dropsonde observations with tropical-cyclone descriptors from the International Best Track Archive for Climate Stewardship (IBTrACS). The framework uses CatBoost, natural-gradient boosting (NGBoost), and a multilayer perceptron (MLP) as base learners, with a random forest (RF) serving as the second-stage nonlinear fusion model. Rather than relying solely on bulk physical parameterization, the framework aims to represent the nonlinear influence of tropical cyclone-related environmental factors on duct strength. Evaluated over 1996–2024, the CNRF attains a test-set R² of 0.791 and a root mean square error (RMSE) of 5.350 M-unit, corresponding to a 23.5% improvement in RMSE over the Naval Postgraduate School (NPS) numerical model. For Hurricane Fiona (2022), the model achieves an RMSE of 6.260 M-unit, and the inclusion of tropical cyclone descriptors improves RMSE by approximately 17.0% relative to a model that excludes tropical cyclone information. The proposed framework facilitates quantitative assessment of extreme-weather-driven duct variability and supports robust design and operation of duct-enabled maritime communication systems. Full article