Researchers have developed a novel approach that integrates complete threedimensional molecular structures with traditional quantitative structureproperty relationship (QSPR) methods using deep learning. By combining molecular descriptors with chargedensity fields from density functional theory, a convolutional neural networkenhanced artificial neural network model significantly improves the prediction of critical temperature and critical pressure for 1359 organic compounds. The model achieves high accuracy (for T_c: R²=0.888, MAPE = 5.03 %; for p_c: R²=0.919, MAPE = 6.37 %), outperforming both conventional QSPR and the widely used JOBACK group contribution method.

A research paper just published in Science China Life Sciences reports that a bacterial effector, delivered via a type IV secretion system, directly disarms a competitor’s type VI secretion system by hijacking its core structural component VgrG1, thereby blocking the competitor's antibacterial defense without killing it — a "win without battle" strategy.

Alterations in the bacterial composition in various anatomical sites of the human body have been associated with tumorigenesis and the progression of multiple cancers; however, divergent results regarding enriched bacteria have been reported across studies of the same disease, indicating cohort-dependent bacterial compositional variance. To move beyond this inconsistency, a research team led by Professor Na Liu from Sun Yat-sen University Cancer Center proposes a fundamental shift in perspective. The team argues that it is the functional repertoire of bacterial communities, rather than their taxonomic identity, that serves as the core driver of tumor progression. The team proposes the unified nomenclature of bacterial functional constituents as “tumor-associated bacterial effectors” (TABEs), categorizing them into six functional classes according to their chemical nature, conserved structural features, and analogous mechanisms of action in regulating host cells. The researchers believe that exploring the mechanisms of TABEs in cancer represents a critical step toward harnessing their biological potential in real-life clinical settings.

Inspecting subwavelength defects in dense nanopatterns is hindered by weak signals and strong background noise. Researchers in China introduced Spatiotemporal Photothermal Modulation Microscopy (SPM²). By leveraging the thermo-optic effect of silicon, they achieved active resonance shifting to amplify defect signals while suppressing backgrounds. This non-destructive technique enables high-sensitivity inspection with microsecond response times. It opens new avenues for wafer inspection and quality control in advanced semiconductor manufacturing.

Programmable optical particle transport based on structured light plays a crucial role in microscale manipulation. Scientists in China have developed a multi-prior physics-enhanced neural network (MPPN-RW) that enables high-fidelity generation of arbitrary optical conveyor belts without training data. This technique allows precise and stable transport of microparticles along complex trajectories, offering new opportunities for optical micromanipulation, targeted delivery, and reconfigurable light-field engineering.

This study demonstrates an experimentally feasible scheme to achieve robust strong coupling between a single quantum dot and a plasmonic nanocavity integrated with a one-dimensional photonic crystal cavity. A Rabi splitting exceeding 170 meV is observed in dark-field scattering spectra. We further demonstrate that the stronger localized electric field within the hybrid cavity not only enhances the coupling strength but also, owing to the more uniform field distribution, reduces the sensitivity of the coupling strength to the quantum dot position within the cavity, thereby improving the uniformity of the device's coupling performance. The robustness of such a strongly coupled system will advance the development of room-temperature quantum devices based on single emitters for potential applications.

Confronting the central challenge of how necroptosis reconciles high sensitivity to stimuli with robustness against intrinsic noise in complex living systems, this study systematically dissect the underlying design principles that govern cell-fate decisions. By integrating biophysical modelling with large-scale topological screening, we resolve the intricate biochemical reaction network into a minimal set of organizing rules. Our analysis identifies the incoherent feedforward loop (IFFL) as the core functional motif driving the observed dynamics. This topology endows the system with a distinctive bell-shaped input–output response, scale invariance, and the capacity to switch precisely between apoptotic and necrotic fates. Beyond elucidating the dynamical logic of cell death signalling, this work reveals, from a physics-informed perspective, a unifying “complexity-to-simplicity” design principle that may underlie the evolutionary construction of sophisticated signalling networks. It further provides a conceptual framework for understanding how dysregulation of cell-fate decisions contributes to pathological processes such as inflammation and tumorigenesis.

When robots perform complex tasks, the pose accuracy of the end-effector is critical. However, errors from individual joints tend to accumulate along the kinematic chain, making it challenging to guarantee high pose precision at the end-effector. To address this issue, this study proposes a virtual-constraints-based end-effector pose compensator (VEPC). The method treats the actual angles of specific joints as known inputs and automatically adjusts the remaining joint angles in real time, effectively eliminating the pose errors of the end-effector caused by the joints. Experimental results demonstrate that the method can reduce the maximum end-effector position error by over 75%. Moreover, the method requires no additional sensors, offering low cost and high compatibility.

A new review shows how PEDOT:PSS, a flexible conductive polymer, enables tactile sensors that mimic human touch. These sensors detect pressure, strain, and temperature, opening doors for smarter healthcare monitoring, robotics, and wearable electronics.

Liver organoids, three-dimensional structures derived from stem cells or hepatic progenitors, have emerged as a transformative technology. Unlike traditional two-dimensional cultures or animal models, organoids faithfully recapitulate the complex architecture and functionality of native liver tissue. This review summarizes recent advancements in liver organoid technology, detailing their development, classification, and key applications.