This retrospective comparative study evaluated whether the adoption of virtual surgical planning (VSP) was associated with changes in adjunctive sliding genioplasty use in patients undergoing bimaxillary orthognathic surgery for facial asymmetry. A total of 117 patients were classified into a model surgery (MS) group (n = 32) and a VSP group (n = 85). Sliding genioplasty frequency was compared between groups, and subgroup analysis was performed according to preoperative menton deviation. In patients with severe menton deviation (≥4 mm), postoperative transverse skeletal and soft-tissue symmetry was additionally assessed using cone-beam computed tomography. Overall, there was slight evidence that adjunctive sliding genioplasty was performed less frequently in the VSP group than in the MS group (18% vs 34%, P = 0.079). In patients with severe menton deviation, the frequency was significantly lower in the VSP group (27% vs 82%, P = 0.003). In this subgroup, both VSP and MS groups showed significant postoperative improvement in transverse facial symmetry, with no significant intergroup differences in postoperative menton deviation (P = 0.17), soft-tissue pogonion deviation (P = 0.099), transverse surgical change at the bony menton (P = 0.52) or soft-tissue pogonion (P = 0.86), or the soft-to-hard tissue movement ratio (P = 0.20). VSP was associated with less frequent adjunctive sliding genioplasty in patients with severe preoperative menton deviation, while postoperative cone-beam computed tomography-based transverse skeletal and soft-tissue symmetry remained comparable between groups. These findings suggest a potential clinical role for VSP in treatment planning for patients with severe facial asymmetry.

This study proposes a 2π solid angle-based heuristic framework for non-coplanar volumetric modulated arc therapy (VMAT) in patients with multiple brain metastases (MBM) to automate treatment angle selection.&#xD;Approach: Fifty-two patients with MBM (2-8 targets) were retrospectively included. A hybrid heuristic framework was constructed: first, a volume-weighted cumulative tumor-skin distance (VW-CTSD) metric was introduced to quantify geometric requirements in the 2π solid angle space and to automatically identify non-coplanar trajectories minimizing normal brain exposure. Subsequently, a blocking factor (BF)-based sub-arc segmentation and collimator-angle selection strategy determined the sub-arc splitting and collimator angles. Using 30 Gy/5 fractions, three plans were generated for each patient: manual expert planning (Pref), VW-CTSD-guided trajectory selection alone (Ptsd), and the proposed combined heuristic strategy (Pcom). Dosimetric metrics and plan parameters were compared.&#xD;Main results: All plans met clinical goals. Ptsd achieved automated trajectory selection with plan quality comparable to Pref. Values are presented as median (P25-P75). Compared with Pref, Pcom improved CI [0.86 (0.85-0.89) to 0.89 (0.85-0.91)] and reduced GI [3.37 (3.17-3.79) to 3.09 (2.74-3.48)]. Pcom also reduced normal brain tissue (NBT) V12Gy, MU, and complexity metric (CM) from 92.45 (62.54-141.46) to 83.57 (50.30-125.10) cc, from 2391 (2144-2527) to 1938 (1845-2085), and from 0.22 (0.21-0.24) to 0.20 (0.19-0.21), respectively. Control-point aperture analysis showed X-axis narrowing [129.75 (110.60-146.35) to 101.91 (83.34-120.20) mm] and Y-axis expansion [98.70 (80.13-105.86) to 119.17 (103.38-133.48) mm] in Pcom. All Pcom-versus-Pref comparisons for these metrics were statistically significant (P < 0.001).&#xD;Significance: The proposed combined heuristic framework reduced NBT dose, GI, MU, and CM, suggesting its potential as a practical automated approach for non-coplanar VMAT angle selection in patients with MBM.

The inverse design of resilient infrastructure materials is hindered by the combinatorial complexity inherent in optimizing stochastic, heterogeneous microstructures. Traditional Euclidean deep learning architectures, such as Convolutional Neural Networks (CNNs), fail to efficiently encode the sparse, non-local topology of disordered porous media, resulting in prohibitive computational redundancy. This work articulates a comprehensive graph-theoretic framework for the predictive modeling and optimization of self-healing cementitious composites. We propose a robust methodology for discretizing volumetric tomography into permutation-invariant heterogeneous graphs, where nodes encode discrete physical entities (aggregates, pores, and microcapsules) and edges represent mechanical and hydraulic connectivity. To capture global fracture dynamics and local transport phenomena, we employ specialized architectures, specifically Pore-GNNs and Long-Short-Edge MeshGraphNets, which function as high-speed surrogate models for multiphysics simulations. Computational validation against high-fidelity numerical solvers confirms the system's efficacy: the proposed architecture achieves approximately 1.1×105-fold acceleration (inference plus preprocessing) in permeability prediction relative to Lattice Boltzmann Method simulations, while maintaining a prediction accuracy of R2 > 0.94 (mean±std over five random splits); the inference-only speedup, which excludes one-time graph-construction preprocessing, is approximately 1.8×106-fold. We note that the end-to-end figure (1.1×105) supersedes the inference-only figure quoted in the original submission metadata. Furthermore, the framework autonomously navigated the high-dimensional design space to isolate a non-intuitive geometric optimum within just 30 high-fidelity oracle evaluations: anisotropic microcapsules (ρc≈2.5) oriented at 45∘ relative to the shear plane. This configuration is predicted to resolve the conflicting physical constraints of the system, maximizing the probability of crack interception for effective healing. These results demonstrate that physics-informed graph learning can effectively decouple prediction accuracy from computational intensity, offering a scalable paradigm for the discovery of advanced functional materials.

The swift vaccine development to combat COVID-19 illustrated the potential for messenger RNA (mRNA) therapeutics to transform drug development. Like mature mRNA, in vitro transcribed mRNA possesses the same elements including a 5' cap, untranslated regions (UTRs), coding sequence and a poly(A) tail. Previous work studying the effects these components have on mRNA translation has primarily utilized highly engineered reporter proteins which exhibit efficient translation and protein stability. With the structural elements of each mRNA differentially affecting their translation, it is imperative to identify the optimal design relevant to the therapeutic protein of interest (POI). To enable POI translation characterization, a split luciferase complementation system was employed. A short peptide tag (HiBiT), which can be fused to either terminus of the POI, associates with its complementary heterodimer (LgBiT) to reconstitute enzymatic activity in the presence of a cell-permeable substrate. To date, split luminescent tagging has been primarily used for high-throughput protein turnover studies. We have previously demonstrated how split luminescent tagging can be employed to enable high-throughput quantification of mRNA translation temporally in cellulo in HEK293 cells constitutively expressing the complementary heterodimer. Here, we further demonstrate the versatility of the assay and detail how this assay can be employed for optimizing in vitro transcription to reduce costs. The assay system can uniquely distinguish alterations in structural components whilst highlighting the effects of coding sequence optimization using non-engineered genes. Additionally, we demonstrate that a 4-fold reduction in 5' cap concentration for in vitro transcription results in equivalent translation in cellulo. These findings illustrate how split luminescent tagging can be easily integrated into the mRNA therapeutic workflow, enabling monitoring of real-time mRNA-driven protein expression dynamics in cellulo thereby offering a versatile method for the advancement of mRNA-based therapeutics.

Unsupervised person re-identification (USL-ReID) typically relies on clustering to generate pseudo-labels, but significant cross-view appearance variations often cause images of the same identity to be split into different clusters. Training on such noisy pseudo-labels severely degrades the learned representations. Therefore, learning robust view-invariant features is paramount. Data augmentation provides a direct way to enhance invariance, yet its trade-offs in USL-ReID remain under-explored: weak augmentations usually preserve identity semantics but lack diversity, whereas strong augmentations provide richer appearance diversity at the cost of partially corrupting identity-consistent semantic cues. To address this challenge, we propose Invariant Representation learning with Progressive Prototype Refinement (IRPP), a unified framework that learns invariant and discriminative features from noisy pseudo-labels. IRPP consists of three synergistic components. First, an Augmented Dual-Contrastive Learning (ADCL) module performs dataset-level prototype-guided invariant learning by contrasting weakly and strongly augmented views against cluster-derived prototypes. Second, an Alignment and Uniformity Learning (AUL) module regularizes the mini-batch-level weak-strong feature geometry, leading to more stable feature distributions under data augmentation. Third, a Progressive Prototype Refinement (PPR) mechanism progressively optimizes cluster centroids into cleaner prototypes, thereby mitigating the influence of noisy pseudo-labels and further strengthening invariant representation learning. This closed-loop design enables prototype-guided contrastive learning, weak-strong regularization, and prototype refinement to mutually reinforce each other. Extensive experiments on standard USL-ReID benchmarks demonstrate that IRPP achieves state-of-the-art performance with a simple and efficient training pipeline. Code is available at https://github.com/Trangle12/IRPP.

The blood-brain barrier (BBB) regulates the entry of chemicals into the central nervous system, and contaminants with high permeability can accumulate in the brain, posing neurotoxicity risks. Given the urgent need for high-throughput neurotoxicity assessment, this study proposes a multifeature fusion framework for identifying and quantifying the BBB permeability to neurotoxic chemicals. BBBProfiler leverages an SE(3)-Transformer to encode the three-dimensional molecular geometry and a deep neural network to encode collision cross-section (CCS) and fingerprint, followed by adaptive integration of modality representations through gating weights. Under the scaffold-split evaluation, BBBProfiler achieves high performance with an area under the receiver operating characteristic curve (AUC) of 93.72% and a recall of 89.19% for classification, and a coefficient of determination (R2) of 0.72 for regression. Its predictive performance was validated using a BBB bioassay (11 chemicals) and an external data set (2023 chemicals), in which the model achieved 94.42% AUC. Substructure mask explanation analysis revealed CCS as an influential feature that correlated well with BBB permeability to chemicals (R2 = 0.84). Model predictions were supplemented with natural-language explanations generated from a large language model. BBBProfiler is deployed on a publicly accessible platform (https://www.ai4environ.cn/BBBProfiler) to facilitate the development of new approach methodologies for neurotoxicity evaluation.

Two-dimensional transition metal dichalcogenide (TMD) interfaces offer a versatile platform for studying quantum phenomena and developing device functionalities. When distinct TMD monolayers are stacked vertically or laterally stitched, their interfaces can exhibit unique electronic band alignments, giving rise to long-lived interlayer excitons, charge transfer effects, and moiré superlattices with correlated states. Here, we demonstrate that the interface between a large variety of two different epitaxially grown TMD monolayers controls the intensity and sign of the Rashba spin splitting, which is probed using THz spintronic emission. Optimized TMD heterobilayers, such as HfSe2/PtSe2, show enhanced THz emission that surpasses the spin-to-charge conversion efficiency of bulk TMDs, confirming the presence of Rashba states with large spin splitting at the interface. By combining spin- and angle-resolved photoemission spectroscopy with density functional theory, we reveal that the electronic hybridization between the two different TMD monolayers gives rise to extended in-gap states with strong Rashba spin-orbit coupling. The choice of TMD layers enables to engineer the sign and strength of spin-to-charge conversion in van der Waals heterobilayers, enabling to build efficient and tunable THz spintronic emitters.

Glecaprevir/Pibrentasvir (GLE/PIB) is an interferon-free direct-acting antiviral (DAA) regimen indicated for the treatment of chronic hepatitis C virus (HCV) infection. However, data on the pharmacokinetics and stability of crushed or split tablets are lacking, complicating administration in patients with dysphagia requiring percutaneous endoscopic gastrostomy (PEG). Additionally, no published studies have evaluated the safety and efficacy of DAAs in combination with Risdiplam, the first oral therapy developed for spinal muscular atrophy (SMA). We report a case of a 14-year-old dysphagic patient with SMA, receiving Risdiplam, who was successfully treated with crushed GLE/PIB administrated via PEG. The patient achieved a sustained virologic response (SVR) 12 weeks after the end of therapy, without significant adverse effects, and no drug interactions were observed during the course of treatment.

Pod shattering, in which mature pods split along their dorsal and ventral sutures, has been widely reported in legume crops due to yield losses. However, the pod-shattering behavior of native legumes that have adapted to decreasing precipitation remains poorly understood. In this study, pod-shattering rates, histology, pod wall components, hydrolase activity and transcriptome of the pod ventral sutures were measured in 6 Caragana species adapted to a wide range of mean annual precipitation (MAP) conditions and growing in a common garden. Species from low-MAP environments presented higher pod-shattering rates, which were associated with a more pronounced abscission zone (AZ), more pod wall torsion laps, lower contents of cellulose and pectin, and increased activities and gene expression levels of polygalacturonase (PG) and cellulase (CE). In contrast, species from high-MAP environments presented lower pod-shattering rates, which were associated with the absence of an AZ, lower pod wall torsion laps, higher cellulose and pectin contents, and reduced activities and gene expression levels of PG and CE. This study hypothesizes that pod-shattering synchronizes seed release to match rainfall patterns, offering a new perspective for understanding the natural distribution of legumes adapted to arid environments.