Browse Articles

Encapsulated droplet array for intensified micromole-scale high-throughput reaction screening

Risk-sensitive collaborative parameter tuning via calibrated deep surrogates for rare-earth electrolysis energy efficiency

Epidemiological and bioinformatics analyses of air pollution and genetic susceptibility in aortic stenosis risk

Spontaneous perceptual reversals are accompanied by systematic changes in pupil size but not respiration phase

Abstract Humans tend to align their respiration with important events, in line with the notion that respiration serves as a tool to allocate (neuro-)physiological resources. This respiration alignment is consistently reported in laboratory studies relying on sequences of temporally structured, stimulus-driven experimental trials. Whether this alignment also occurs in the absence of external cues - such as during spontaneous changes of perception - remains unknown. Notably, pupil size is known to change around spontaneous changes in perception, and pupil size covaries with respiration phase, suggesting that respiration may also align to spontaneous changes in perception. We here investigate the three-fold relations between respiration, pupil size, and spontaneous perceptual changes in a paradigm where participants ( N  = 20) reported spontaneous perceptual reversals of an ambiguous Necker cube. We confirmed the biphasic modulation of pupil size around reversals (constriction followed by dilation) and the covariation of pupil size with respiration phase. However, the alignment of respiration phase to perceptual reversals was much weaker than previously reported for stimulus-driven experimental trials. These findings suggest that changes in perception alone, whether exogenously or endogenously driven, might not be the key driving force of respiration alignment, but the temporal predictability of external events likely facilitates this alignment.

Dynamics of sociopolitical polarization and effects of misperception-correcting information around the 2022 Brazilian elections

Abstract Affective polarization is little-studied in systems that present ambiguity in relevant political groups. We assess dynamic patterns in intergroup liking and perceived intelligence over 5 waves of a pre-registered survey with 4 panel waves (N ≃ 2,000-3,000) conducted before, during, and after the 2022 Brazilian elections. Given the nature of Brazil’s political system, we document horizontal affective polarization of combinations of political ingroups and outgroups. These include left-right self-placed position-holders, negative partisans, and supporters of the most important presidential candidates, “Lula” and “Bolsonaro”. Categorized by first-round vote intention, the latter grouping (lulistas vs. bolsonaristas) shows particularly intense affective polarization before the election campaign, which remains stable through the elections, thereafter fading approximately symmetrically. We also demonstrate that experimental treatment conditions correcting misperceptions of group stances on single, controversial policies reduce affective polarization by up to 6.66 percentage points (liking thermometer), and, in some instances, also shift policy support.

Federated autoencoder-based clinical decision framework with hybrid class balancing

Abstract Despite diagnosis accuracy has been much improved by depending more on deep learning for disease classification, it raises serious concerns about patient data privacy, security, and scalability. Conventional centralized deep learning approaches are vulnerable to data leaks and non-compliance with such privacy rules as GDPR and HIPAA since they rely on the aggregation of sensitive medical records. Our work presents a privacy-preserving federated learning architecture enabling cooperative model training among many healthcare institutions without exposing raw patient data, hence addressing these issues. Combining autoencoder-driven hierarchical feature extraction, the proposed method improves classification performance and guarantees low information loss. Moreover applied is a hybrid class-balancing mechanism integrating generative augmentation techniques with Synthetic Minority Over-Sampling Technique (SMote) to eliminate bias in unbalanced illness datasets, so raising sensitivity for minority-class scenarios. Maintaining computational efficiency in federated systems, experimental evaluation demonstrates that the proposed model achieves accuracy of 92.5% outperforming both classic CNN-based (87.2%) and LSTM-based (89.1%) models. Strong convergence even in non-IID distributed medical data is promised by adaptive federated averaging. Furthermore, the proposed method resists adversarial attacks, therefore enhancing the security in the surrounding practical areas. This work decreases the distance between high-accurate disease categorization and privacy-preserving artificial intelligence by offering the basis for scalable, distributed, secure medical intelligence systems. The findings help federated medical artificial intelligence to grow by proving its ability to change healthcare diagnostics while keeping regulatory compliance and data security.In this study, the Synthetic Minority Over-Sampling Technique (SMOTE) is incorporated to address the class imbalance present in the distributed medical datasets. SMOTE generates new synthetic samples for minority classes by interpolating between existing minority instances, thereby preventing model bias toward majority classes and improving classification robustness across federated clients. By enhancing minority-class representation prior to federated aggregation, SMOTE ensures that the global model learns more discriminative and balanced feature patterns.

Broadly quantifying SSP elements for 188 countries to 2150 in International Futures

Benefits of building information modeling (BIM) for construction projects success in Pakistan

Manipulating electrolyte solvation structures to build high-voltage concentration batteries for efficient energy storage

Validation of the modified transplant symptom occurrence and symptom distress scale and exploring symptom burden in Brazilian kidney transplant recipients

Crossed laser phase plates for transmission electron microscopy

Abstract A phase plate has long been sought in transmission electron microscopy (TEM) to maximize the image contrast of weakly-scattering objects like biomolecules. The laser phase plate (LPP) has recently demonstrated that an amplified, focused laser standing wave reliably phase shifts the electron beam, achieving phase-contrast TEM. Building on the single-beam LPP, here we introduce the crossed laser phase plate (XLPP): two laser standing waves which intersect in the diffraction plane. We present a theoretical model for the XLPP inside the microscope and show that, relative to the original LPP, it increases information transfer at low spatial frequencies while suppressing ghost images formed by Kapitza-Dirac diffraction. We also present a simple acquisition scheme, enabled by the XLPP, which further suppresses ghosts. Finally, we discuss practical considerations of XLPP design and show experimental results from a prototype. The results of this study chart the course for future developments of LPP hardware.

Factors associated with post-traumatic growth in severe COVID-19 patients discharged from intensive care: the influence of near-death experiences

Chiral emission from chirotopic nanostructures

Modeling anisotropic preference manifolds for robust graph-based fashion recommendation

Optimal blood pressure target in patients with uncomplicated hypertension: a target trial emulation study

Abstract The limited real-world evidence on the clinical benefits of intensive blood pressure (BP) management demonstrated in randomised controlled trials has led to its poor adoption in primary care settings. Here we conducted a target trial emulation study on 118,271 patients with uncomplicated hypertension to evaluate the effectiveness and safety managed by intensive (below 130/80 mmHg) compared to standard (130-140/80-90 mmHg) BP targets using a territory-wide public healthcare database in Hong Kong. Patients in the intensive blood pressure target group were observed to incur a lower risk of hypertension related complications and all-cause mortality with no significant increased risk of the serious adverse events reported. The findings of this study provide evidence on the clinical benefits of an intensive BP management in real-world settings, supporting the adoption of a BP target of less than 130/80 mmHg in clinical guidelines for the treatment of adult patients with hypertension in primary care.

Optical response in six-level 85Rb and 87Rb atomic media

Adhesion-derived condensates control component availability to regulate adhesion dynamics

Abstract Integrin adhesion complexes mediate dynamic cell-matrix interactions to regulate cell adhesion, migration and invasion. While increasing evidence suggests that adhesion components can undergo liquid-liquid phase separation, the functional relevance of adhesion-derived condensates remains elusive. Here, we demonstrate that tensin 1 (TNS1), a multidomain adaptor protein linking active integrins with the actin cytoskeleton, forms condensates in cells to regulate adhesion component availability. We show that endogenous TNS1 condensates are formed upon focal adhesion disassembly or limited integrin-extracellular matrix engagement, acting as reversible reservoirs for unphosphorylated adhesion proteins. We identify the TNS1 intrinsically disordered region as governing TNS1 condensation and condensate biophysical properties. We further demonstrate a negative regulatory role for phosphorylation on condensate assembly upon stress-responsive kinase activation. Finally, we confirm the functional effects of TNS1 condensation on adhesion dynamics and cell migration, identifying a feedback loop acutely regulating kinase-mediated adhesion turnover upon integrin activation.

Deep learning based on multiparametric MRI for differentiating benign from malignant PI-RADS 3 prostate lesions: a dual-center study

Abstract To address the diagnostic challenge of indeterminate PI-RADS 3 prostate lesions, we developed a non-invasive deep learning radiomics model using multiparametric MRI. This dual-center retrospective study included 382 patients. Lesions on T2WI, DWI, and ADC were segmented and standardized. Deep learning features were extracted via ResNet101, and radiomics features were selected using minimum Redundancy Maximum Relevance (mRMR) and Least Absolute Shrinkage and Selection Operator (LASSO). We built and compared three models: clinical, deep-learning radiomics signature (DLRS), and a combined model. The combined model achieved the highest AUCs (0.951, 0.964, 0.925), significantly outperforming the clinical model ( p  < 0.05). The DLRS model also showed high performance (AUCs: 0.946, 0.932, 0.918). The nomogram was well-calibrated, and DCA indicated clinical utility. This mpMRI-based deep learning model accurately differentiates malignant from benign PI-RADS 3 lesions, offering a valuable non-invasive tool to aid clinical decisions and reduce unnecessary biopsies.

Identification of a mechanism-based binding mode for a histone deacetylase 6 inhibitor

Abstract Histone deacetylase 6 (HDAC6) is a cytoplasmic enzyme that deacetylates non-histone substrates such as α-tubulin and cortactin. HDAC6 contains two catalytic domains, each containing a catalytic zinc ion, and a zinc-finger ubiquitin-binding domain. We have discovered BAS-2, a selective HDAC6 inhibitor with an isothiouronium core and no obvious zinc-binding group. To define its mechanism, we combine X-ray crystallography, structure-activity-relationships, molecular modeling and mutagenesis. BAS-2 potently inhibits human HDAC6 but it does not inhibit zebrafish HDAC6. Computational modeling highlighted Asp567 in human HDAC6 as critical for BAS-2 recognition and mutational analyses confirmed this. The corresponding zebrafish residue is Asn530 and the crystal structure of the N530D variant zHDAC6 revealed binding of a BAS-2–derived mercaptoacetamide that engages the catalytic zinc via strong thiolate–zinc coordination. Leveraging the orientation of BAS-2 binding, we designed a BAS-2–based proteolysis targeting chimera that induced proteasome-dependent HDAC6 degradation in cells, verified by global proteomics. Collectively, these insights clarify species selectivity and demonstrate that BAS-2 acts as a selective, mechanism-based inhibitor of human HDAC6. These discoveries will aid the development of the next generation of selective HDAC6 inhibitors and degraders.

Mixed channel traffic grooming with crosstalk awareness and shared backup path protection in IP over multicore fibre elastic optical networks

Abstract Multicore fibre elastic optical networks (MCF-EONs) offer a promising route to overcome the capacity limit of single-core systems, but practical deployment is hindered by inter-core crosstalk and by the poor resource efficiency of dedicated protection. Here we address mixed channel traffic grooming in IP over MCF-EONs protected by shared backup paths, in which working and protection flows can be aggregated onto common optical channels subject to crosstalk constraints. We develop a double auxiliary-graph routing, core and spectrum assignment framework (Double-AG-RCSA) that jointly coordinates IP-layer grooming with crosstalk aware optical-layer provisioning. Simulations on two backbone topologies equipped with 7-core and 19-core fibres show that the proposed framework reduces the maximum used frequency-slot index by up to 33%, lowers the average inter-core crosstalk by about 6 dB, and decreases transponder usage by 28 to 35% relative to representative baselines, while eliminating crosstalk-threshold violations. The results demonstrate that cross-layer coordination of grooming, crosstalk aware provisioning and shared backup protection improves survivability, spectrum efficiency and signal quality in multi-layer optical networks. The framework is compatible with existing control planes and can be extended to dynamic traffic and larger topologies.