Search

A modularized open-source pipeline for invasive brain signal decoding bridges the gap between closed-loop neuromodulation and clinical brain–computer interface approaches in a large patient cohort.

The authors report the implementation of a Transformer-based model on the same architecture used in Large Language Models in a 14nm analog AI accelerator with 35 million Phase Change Memory devices, which achieves near iso-accuracy despite hardware imperfections and noise.

Genome-wide association meta-analyses of attention-deficit/hyperactivity disorder symptom measures and clinical diagnoses identify new risk loci, highlight putative effector genes and yield improved polygenic risk scores.

Metasurfaces can solve Fredholm integral equations of the second kind for free-space radiation at optical wavelengths. To this end, an inverse-designed metagrating is coupled to a semitransparent mirror providing feedback to perform an analogue version of the Neumann series.

By using a resonant sensor to couple two radio-frequency parametric oscillators behaving as Ising spins, a passive wireless device can implement programmable temperature threshold sensing.

In quantum information the majorization conjecture states that the minimum amount of disorder at the output of a quantum Guassian channel is produced by coherent input states, but its proof has remained elusive. Now, Mari et al.solve this longstanding problem and highlight some of its implications.

Methods to train physical neural networks, such as backpropagation-based and backpropagation-free approaches, are explored to allow scaling up of artificial intelligence models far beyond present small-scale laboratory demonstrations, potentially enhancing computational efficiency.

Genetically encoded sensors have been developed and become versatile tools for imaging serotonin dynamics. Here, authors developed a family of serotonin (5-HT) sensors (sDarken), three variants with different affinities for 5-HT enable high spatiotemporal resolution of 5-HT dynamics.

The mechanisms underlying the zebrafish spinal cord’s impressive innate ability to regenerate after traumatic insults remain largely unknown. Here, the authors show that spinal neurons enact an essential neuron-to-neuron neuroprotective strategy to safeguard cell survival and support the zebrafish’s regeneration process.

The underpinnings of a supposed basal ganglia pathway imbalance in dystonia are unknown. Here, the authors unveil exaggerated low frequency coupling in the direct striato-pallidal pathway to reflect dystonic symptoms, which could potentially be used as target for neuromodulation strategies.

Device-level complexity represents a big shortcoming for the hardware realization of analogue memory-based deep neural networks. Mackin et al. report a generalized computational framework, translating software-trained weights into analogue hardware weights, to minimise inference accuracy degradation.

The mechanisms stimulating adult neurogenesis are unclear. Here, the authors show the contribution of cholinergic and GABAergic signalling within the locomotor network to spinal cord neurogenesis during homeostasis and regeneration, showing neurogenesis depends on circuit activity in the adult zebrafish.

Using well-calibrated statistical methods the authors jointly analyze Undiagnosed Diseases Network genomes, identifying known and novel disease genes. Software is publicly available to support future cross-cohort rare disease discovery efforts.

Dopaminergic medication and deep brain stimulation exert a shared therapeutic modulation of cortex-subthalamic nucleus network activity in Parkinson’s disease, through a suppression of high beta synchrony mediated by the hyperdirect pathway.

Large absorption of high-index semiconductors has hindered the application of all dielectric nanostructures in the visible range. Here, the authors present bandgap-engineered hydrogenated amorphous Si nanoparticles with Q-factors up to 100 and their integration with photochromic molecules as tunable meta-atoms.

Dynamical quantum phase transitions can be observed when time is treated as a control parameter in non-equilibrium quantum systems. The authors show that quenching the qubits-oscillator coupling in a dissipative two-qubit system leads to different transitions depending on interactions and entanglement, with promising applications in quantum sensing and metrology.

Thermal crosstalk is a major source of low accuracy when training Photonic Neural Networks. In this study, the authors present a model for Mach-Zehnder interferometer-based devices that accurately predicts thermal effects and accounts for fabrication variations, providing a reliable tool for photonic computing simulation and design.

Here, the authors present and characterise a collection of human gut bacteria including novel taxa associated with health conditions and a large diversity of plasmids. All isolates, their genomes and metadata are publicly available, facilitating research by others (www.hibc.rwth-aachen.de).

Transport in quantum systems is complex and can be suppressed by coherent superposition of the involved states. Here, the authors find all-electronic suppression of transport in a carbon nanotube originating from coherent population trapping and give criteria for the presence of such a dark state.

Although the heart has its own nervous system, its organization and functionality remain largely unknown. Here, the authors reveal the molecular, chemical, and functional diversity of neurons within the intracardiac nervous system and their role in controlling the heart’s rhythm in the zebrafish.

Arrays of silicon nanoneedles are used to generate molecular replicas of live brain tissue for longitudinal spatial lipidomic classification via desorption electrospray ionization mass spectrometry imaging of gliomas and to monitor the responses of the tumours to chemotherapy.

To realize electronic operations beyond the von Neumann bottleneck, a new type of switch that can mimic self-learning is needed. Here, the authors demonstrate all-in-one-place logic and memory operations based on dynamic molecular switch that can emulate brain-like synaptic and Pavlovian response, bringing the field a step closer to molecular-scale hardware.

Analog in-memory computing promises efficient DNN inference acceleration but suffers from nonidealities. Here, hardware-aware training methods are improved so that various larger DNNs of diverse topologies nevertheless achieve iso-accuracy.

Nonlinear optical computations have been essential yet challenging for developing optical neural networks with appreciable expressivity. In this paper, light scattering is combined with optical nonlinearity to empower a high-performance, large-scale nonlinear photonic neural system.

A large-scale CRISPR-mediated deep mutational scanning approach is used to interrogate the function of mutations in the endogenous locus of TP53 mapping to the DNA-binding domain.

Unlike most eukaryotes, Luzula sylvatica assembles holocentromeres along the entire chromosome length. Here, the authors combine genome assembly, epigenetic and cytogenetic analyses to reveal holocentricity involves the fusion of small monocentric chromosomes and the expansion of CENH3-interacting satellite DNA repeats.

Using a high-quality chromosome-scale assembly of the faba bean genome, the genetic basis of seed size and hilum colour is explored.

Cross-shore channels can counterintuitively form at coasts dominated by alongshore currents. Sharp changes in bed elevation around the mean sea level lead to nearly 90-degree intersection angles between the channel and the shoreline.

Nanophotonic light sources with programmable emission spectrum are important building blocks for integrated photonics, sensing and optical computing. Here the authors tune the complex laser spectrum of a network laser achieving selective lasing of a single, two or more modes.

In hepatocellular carcinoma driven by non-alcoholic steatohepatitis, aberrant T cell activation and impaired immune surveillance seem to make hepatocellular carcinoma less responsive to anti-PD1 or anti-PDL1 immunotherapy.

Genomic and molecular analyses of Clunio marinus timing strains suggest that modulation of alternative splicing of Ca2+/calmodulin-dependent kinase II represents a mechanism for evolutionary adaptation of circadian timing.

Common bean is an evolutionary model for studying adaptive diversity in legumes. Here, the authors present the common bean pangenome based on five high-quality genomes and whole-genome reads of 339 wild and domesticated genotypes, and reveal adaptive gene loss during expansion and domestication.

Using functional magnetic resonance imaging, Setti et al. show that the superior temporal cortex is synchronized across auditory and visual presentation of the same narrative, even in sensory-deprived individuals who lack any audiovisual experience.

The exciton–phonon coupling (EXPC) affects the opto-electronic properties of atomically thin semiconductors. Here, the authors develop two-dimensional micro-spectroscopy to determine the EXPC of monolayer MoSe₂.

The authors demonstrate broadband and efficient radio-transparent antennas based on cloaking technology. Their features are well suited for modern communication systems, as closely spaced antennas need to be integrated with minimal interference.

Gioia, Tavella et al. show that severe acute respiratory syndrome coronavirus 2 causes DNA damage through CHK1 degradation and impairs 53BP1 recruitment to DNA lesions. The induced DNA damage is associated with expression of pro-inflammatory cytokines and senescence markers.

Common bean has two distinct domestication centers in Mesoamerica and in the Andes. The authors show that the Andean is the first gene pool successfully introduced in Europe and identify signature of pervasive introgression among gene pools and of selection for flowering underlying adaptation.

Single-particle resolved measurements in an ultracold-atom experiment reveal an intuitive picture of quantum correlations, providing strong constraints on the full density matrix of the two-particle system in the presence of interactions.

There are a number of non-trivial integrable models in one-dimension, making them an attractive starting point for studying quantum dynamics. Biella et al. study transport between two semi-infinite solvable models and show that a slowly-relaxing region forms around the integrability-breaking junction.

Homologous recombination (HR) gene mutations are thought to be synthetic lethal with DNA polymerase theta (Polθ) inhibition. Here, the authors reveal that Polθ addiction is determined by the functional impact of gene mutations on DNA end resection activity.

Eberhard et al. show that SEMA3A regulates liver sinusoidal endothelial cell fenestrations by signaling through NRP1 and LIMK1, revealing a pathway that connects hyperlipidemia to the development of steatotic liver disease.

The modes of random lasers, by their very nature, are difficult to design and control. Here, Gaio, Saxena et al. demonstrate a laser based on a nanophotonic network, instead of random scatterers, which allows tailoring of the optical properties and the lasing output via the network topology.

All components of the proton–proton nuclear fusion chain, in which hydrogen is converted into helium in the Sun, are described, with several implications for fundamental solar and particle physics.

Plasma extracellular vesicles contain quantifiable amounts of TDP-43 and full-length tau, allowing the accurate assessment of pathology in frontotemporal dementia, frontotemporal dementia spectrum disorders and amyotrophic lateral sclerosis.