Browse Articles

Single-image inference of clathrin-mediated endocytosis dynamics via deep learning

Clathrin-mediated endocytosis (CME) is a vital cellular process that exhibits spatial and temporal heterogeneity in its dynamics, traditionally studied through labor-intensive time-lapse microscopy and single particle tracking. To overcome the limitations posed by phototoxicity, temporal undersampling, and computational complexity, we introduce a deep learning framework that infers CME dynamics from single fluorescence images. Using a modified U-Net architecture, our model predicts spatial maps of the standard deviation (SD) of clathrin coat growth rates—an established metric of CME activity—directly from static frames of AP2-eGFP–labeled cells. The network was trained on paired image data and SD maps derived from experimentally tracked endocytic events. The model accurately recapitulates dynamic features such as front–rear asymmetry in migrating cells and responses to membrane tension alterations, demonstrating strong agreement with traditional time-lapse-derived metrics. This approach eliminates the need for trajectory reconstruction or prolonged imaging, enabling real-time, non-invasive assessment of endocytic dynamics across diverse biological contexts. Our results highlight the potential of deep learning to extract dynamic biophysical information from static imaging data and establish a scalable methodology for probing CME and related subcellular processes.

Kinetochore-centrosome feedback linking CENP-E and Aurora kinases controls chromosome congression

Abstract Chromosome congression is crucial for accurate cell division, with key roles played by kinetochore components, the molecular motor CENP-E/kinesin-7, and Aurora B kinase. However, Aurora B kinase can both inhibit and promote congression, suggesting the presence of a larger signaling network. Our study demonstrates that centrosomes inhibit congression initiation when CENP-E is inactive by regulating the activity of kinetochore components. Depletion of centrioles via Plk4 kinase inhibition allows chromosomes near acentriolar poles to initiate congression independently of CENP-E. At centriolar poles, high Aurora A kinase enhances Aurora B activity, increasing phosphorylation of microtubule-binding proteins at kinetochores and preventing stable microtubule attachments in the absence of CENP-E. Conversely, inhibition of Aurora A or expression of a dephosphorylatable mutant of the kinetochore microtubule-binding protein Hec1 enables congression initiation without CENP-E. We propose a negative feedback mechanism involving Aurora kinases and CENP-E that regulates the timing of chromosome movement by modulating kinetochore–microtubule attachments and fibrous corona expansion, with the Aurora A activity gradient providing critical spatial cues for the network’s function.

Implementing ensemble of deep learning model with optimization techniques for human activity recognition to assist individuals with disabilities

On decoherence in surface hopping: The nonadiabaticity threshold

This study presents a strategy to efficiently and safely account for decoherence in the fewest switches surface hopping method. Standard decoherence corrections often lead to overly strong coherence suppression. A simple and general solution to this problem is to restrict decoherence to regions of low nonadiabaticity, measured by the dimensionless Massey parameter. The same threshold values are suitable for a variety of systems, regardless of their size and absolute energy scale. When restricted to uncoupled regions, a Gaussian overlap decoherence correction consistently leads to more accurate populations than using no correction. The article also examines under what circumstances it is appropriate to decohere instantaneously.

Microbial oxidation significantly reduces methane export from global groundwaters

Methane is ubiquitous in groundwater, and its release to surface environments through pumping, discharge, or diffusion is an emerging environmental concern. Microbial oxidation consumes methane and mitigates its release, but quantitative constraints in groundwater remain unknown. Using ultra-low-level 14 C-labeling, we estimate in situ microbial methane oxidation rates in shallow carbonate and sandy aquifers from central and northern Germany with methane concentrations spanning 5 orders of magnitude, from 0.15 ± 0.04 to 36,250 ± 1,390 µg L −1 . Oxidation rates ranged from 0.001 ± 0.0003 to 74.28 ± 46.94 µgC L −1 d −1 and were highly correlated with groundwater methane concentrations. Oxidation-based methane turnover was rapid at low methane concentrations, with complete consumption requiring days to weeks. In contrast, microbial oxidation at high methane sites required months to decades for complete methane turnover, indicating the potential for unconsumed methane to leak into local streams or wetlands. High oxidation rates were associated with gammaproteobacterial methanotrophs that typically thrive in suboxic conditions and anaerobic methane-oxidizing archaea, while uncultivated methanotrophs of the Methylomirabilota and Verrucomicrobiota dominated low-rate sites. Based on globally distributed groundwater methane concentration data, we extrapolated the strong observed correlation between methane concentrations and oxidation rates to global groundwater volumes, estimating that microbial oxidation removes ~66% of groundwater methane globally, equivalent to 167 to 778 Tg CH 4 y −1 . This highlights the groundwater microbiome as a crucial subsurface methane filter that reduces methane release to surface waters, soils, and the atmosphere.

Directed self-assembly of chiral liquid crystals into biomimetic bouligand structures in thin film

An exploratory binding study of molnupiravir efficacy against emerging Omicron SARS-CoV-2 variants

Reweighting estimator for <i>ab initio</i> path integral Monte Carlo simulations of fictitious identical particles

The fermion sign problem constitutes one of the most fundamental obstacles in quantum many-body theory. Recently, it has been suggested to circumvent the sign problem by carrying out path integral simulations with a fictitious quantum statistics variable ξ, which allows for a smooth interpolation between the bosonic and fermionic limits [Xiong and Xiong, J. Chem. Phys.157, 094112 (2022)]. This ξ-extrapolation method has subsequently been applied to a variety of systems and has facilitated the analysis of an x-ray scattering measurement taken at the National Ignition Facility with unprecedented accuracy [Dornheim et al., Nat. Commun. 16, 5103 (2025)]. Yet, it comes at the cost of performing an additional 10–20 simulations, which, in combination with the required small error bars, can pose a serious practical limitation. Here, we remove this bottleneck by presenting a new reweighting estimator, which allows the study of the full ξ-dependence from a single path integral Monte Carlo (PIMC) simulation. This is demonstrated for various observables of the uniform electron gas and also warm dense beryllium. We expect our study to be useful for future PIMC simulations of Fermi systems, including ultracold atoms, electrons in quantum dots, and warm dense quantum plasmas.

Electrical and thermal conductivity of Earth’s iron-enriched basal magma ocean

The Earth’s earliest magnetic field may have originated in a basal magma ocean, a layer of silicate melt surrounding the core that could have persisted for billions of years. Recent studies show that the electrical conductivity of liquid with a bulk silicate Earth composition exceeds 10 4 S/m at basal magma ocean conditions, potentially surpassing the threshold for dynamo activity. Over most of its history however, the basal magma ocean is more enriched in iron than the bulk silicate Earth, due to iron’s incompatibility in the mineral assemblages of the lower mantle. Using ab-initio molecular dynamics calculations, we examine how iron content affects the silicate dynamo hypothesis. We investigate how the electrical conductivity of silicate liquid changes with iron enrichment, at pressures and temperatures relevant for Earth’s basal magma ocean. We also compute the electronic contribution to the thermal conductivity, to evaluate convective instability of basal magma oceans. Finally, we apply our results to model the thermal and magnetic evolution of Earth’s basal magma ocean over time.

Porous membranes enable selective and stable zero-gap acidic CO2 electrolysers

Mechanical and microstructural enhancement of clayey soils through recycled glass powder stabilization

Viscosity, breakdown of Stokes–Einstein relation, and dynamical heterogeneity in supercooled liquid Ge2Sb2Te5 from simulations with a neural network potential

Phase change materials are exploited in non-volatile electronic memories and photonic devices that rely on a fast and reversible transformation between the amorphous and crystalline phases upon heating. Recrystallization of the amorphous phase under the operation conditions of the memories occurs in the supercooled liquid phase above the glass transition temperature Tg. The dynamics of the supercooled liquid is thus of great relevance for the operation of the devices and, close to Tg, also for the structural relaxations of the glass that affect the performances of the memories. Information on the atomic dynamics is provided by the diffusion coefficient (D) and by the viscosity (η), which are, however, both difficult to be measured experimentally under the operation conditions of the devices due to fast crystallization. In this work, we leverage a machine learning interatomic potential for the flagship phase change compound Ge2Sb2Te5 to compute η, D, and the α-relaxation time in a wide temperature range from 1200 K to about 100 K above Tg. Large-scale molecular dynamics simulations allowed quantifying the fragility of the liquid and the occurrence of a breakdown of the Stokes–Einstein relation between η and D in the supercooled phase. Isoconfigurational analysis provided a visualization of the emergence of dynamical heterogeneities responsible for the breakdown of the Stokes–Einstein relation. The analysis revealed that the regions of most mobile atoms are related to the presence of Ge atoms with particular local environments.

Fluorine-free gel polymer electrolyte for lithium oxide-rich solid electrolyte interphase and stable Li metal batteries

Impact of habitat associations on saproxylic beetle assemblages and their damage severity

Abstract Saproxylic beetles, as primary decomposers in forest ecosystems, play a crucial role in the decomposition of dead wood. However, there is a significant gap in understanding the extent of assemblages and damage caused by these insects, which is essential for managing the quality and utilization of dead wood resources in natural forests. This study employed the Bevan damage classification system to quantify the severity of saproxylic beetle damage to fallen trees, focusing on the boreal forest in the Green zone of Ulaanbaatar, the capital of Mongolia. A factorial design was used to assess the influence of forest landscape (north vs. south mixed forest), tree species (Siberian spruce Picea obovata and Siberian pine Pinus sibirica), and decay class (1–4) on beetle damage indices, abundance and feeding guilds (cambium consumers, wood borers, predators, parasitoids, and detritivores). Our findings reveal that decay class significantly affects beetle abundance and damage severity with early stages showing the highest values. Cambium consumers and wood borers were more abundant in decay class 1 (DC1) for downed spruce, with Ips typographus (24.7%) and Tetropium castaneum (15%) causing the most damage. For the Siberian pine, Monochamus galloprovincialis (9.8%) and Judolia sexmaculata (13.3%) were the most damaging in DC1 followed by Pityogenes conjunctus (10%). The results suggest that Siberian spruce may be more susceptible to saproxylic beetle damage than the Siberian pine, with structural features such as bark cover branch size and wood moisture playing a critical role, especially in early decay stages. Based on our findings, we recommend decay-stage-specific management approaches, particularly targeting early decay stages (DC1–DC2) where beetle damage is most severe. Practical strategies include early detection of freshly downed trees, bark removal to reduce suitable habitat for cambium consumers, and on-site processing techniques such as bark gouging or mechanical debarking. These methods allow deadwood biomass to be retained in the forest while reducing pest pressure, offering a viable alternative to salvage logging. Such approaches are especially relevant in protected areas, where they can support both pest control and biodiversity conservation objectives. However, given the geographic scope limited to boreal forests of Ulaanbaatar, caution should be exercised in extrapolating these recommendations to other regions without further study.

Hybrid exchange–correlation potential built on the piecewise linearity conditions of both energy and electron density

The piecewise linearity condition is a fundamental concept in density functional theory (DFT) and provides powerful conditions for developing accurate approximations. The piecewise linearity condition of energy has been widely used for developing new approximations, while the piecewise linearity condition of electron density has not received much attention in these developments. In this work, we develop a method for building the exchange–correlation (XC) potential based on the piecewise linearity conditions of both energy and electron density. The XC potential is defined as a linear mixing of the exchange potentials from the exact exchange and the local density approximation. The mixing parameter is spatially dependent and is fully determined based on these two linearity conditions. The numerical tests show that the energy’s linearity condition holds well as the system’s electron number is varied between N and N − 1, leading to reliable predictions of the eigenvalues of the highest occupied molecular orbitals of various molecular systems. The density’s linearity condition is well satisfied near N but is violated near N − 1. Finally, this method is examined with more challenging tests: calculating the Kohn–Sham correlation potentials of He, Be, and H2. The key features of these correlation potentials, such as the shell structures and the barriers at bond middle points, are semi-quantitatively captured by this new method, which demonstrates that the piecewise linearity condition of electron density is a promising condition for developing high-quality approximations in DFT.

Field-resilient supercurrent diode in a multiferroic Josephson junction

Abstract The research on supercurrent diodes has surged rapidly due to their potential applications in electronic circuits at cryogenic temperatures. To unlock this functionality, it is essential to find supercurrent diodes that can work consistently at zero magnetic field and under ubiquitous stray fields generated in electronic circuits. However, a supercurrent diode with robust field tolerance is currently lacking. Here, we demonstrate a field-resilient supercurrent diode by incorporating a 2D multiferroic material into a Josephson junction, and observed a pronounced supercurrent diode effect at zero magnetic field. More importantly, the supercurrent rectification persists over a wide and bipolar magnetic field range beyond industrial standards for field tolerance. By theoretically modeling a multiferroic Josephson junction, we unveil that the interplay between spin-orbit coupling and multiferroicity underlies the unusual field resilience of the observed diode effect. This work introduces multiferroic Josephson junctions as a new field-resilient superconducting device for cryogenic electronics.

Synthesis and characterization of Fe₂O₃ nanoparticles via pulsed laser ablation in liquids: effects of solvent and laser fluence

Vapor–liquid equilibrium of water with the machine-learned ML-BOP model

Over the past few decades, many classical force-fields have been developed to model water. However, capturing the properties of water across its solid, liquid, and vapor phases remains a challenge. The coarse-grained machine-learned bond order potential (ML-BOP) model accurately reproduces the structural and thermodynamic properties of liquid water in both stable and supercooled states, as well as the thermodynamics of ice-water equilibrium and polyamorphism, comparable to all-atom models TIP4P/2005 and TIP4P/Ice, but with nearly 100 times lower computational cost. In this study, we evaluate the ability of ML-BOP to describe vapor–liquid coexistence properties of water, despite its development excluding any such training data. We find that ML-BOP underestimates the surface tension at ambient conditions and its slope of temperature dependence, a trend common among coarse-grained models lacking explicit hydrogen atoms. Nevertheless, ML-BOP accurately reproduces vapor–liquid coexistence densities and predicts the critical point (Tc = 653.27 ± 3.0 K and ρc = 0.328 ± 0.004 g cm−3) in excellent agreement with experiment and comparable to TIP4P/2005. ML-BOP outperforms the widely used mW model in reproducing vapor–liquid coexistence properties of water. ML-BOP also captures the high-temperature inflection in the surface tension curve and the specific surface entropy anomaly, predicting the temperature of maximum surface entropy closer to experiment than TIP4P/2005. Furthermore, we investigate the Guldberg and Eötvös empirical relationships in ML-BOP, demonstrating quantitative predictions of boiling-critical temperature scaling and enthalpy of vaporization. Overall, ML-BOP offers a promising balance of accuracy and efficiency, making it the most capable coarse-grained water model currently available for simulating water across various regimes.

Still no evidence for an environmentally responsive epigenetic clock in an insect

The role of phosphorus in the solid electrolyte interphase of argyrodite solid electrolytes

Abstract The solid electrolyte interphase that forms on Li6PS5Cl argyrodite solid electrolytes has been reported to continually grow through a diffusion-controlled process, yet this process is not fully understood. Here, we use a combination of electrochemical and X-ray photoelectron spectroscopy techniques to elucidate the role of phosphorus in this growth mechanism. We uncover how Li6PS5Cl can decompose at potentials well above the full reduction to Li3P, forming partially lithiated phosphorus species, Li x P. We provide evidence of a gradient of Li x P species throughout the solid electrolyte interphase and propose a growth mechanism in which the rate-determining step is the diffusion of lithium through Li x P. We predict continuous solid electrolyte interphase growth as long as metallic lithium is present and a Li x P percolation pathway exists, highlighting the importance of understanding and engineering solid electrolyte interphase composition and nanostructure in solid-state batteries. We believe that this growth mechanism would apply to any solid electrolyte interphase that can contain partially lithiated phosphorus, or potentially any lithium alloy.