Browse Articles
Discover research articles across all indexed journals
Foundation neural-networks quantum states as a unified Ansatz for multiple hamiltonians
Abstract Foundation models are highly versatile neural-network architectures capable of processing different data types, such as text and images, and generalizing across various tasks like classification and generation. Inspired by this success, we propose Foundation Neural-Network Quantum States (FNQS) as an integrated paradigm for studying quantum many-body systems. FNQS leverage key principles of foundation models to define variational wave functions based on a single, versatile architecture that processes multimodal inputs, including spin configurations and Hamiltonian physical couplings. Unlike specialized architectures tailored for individual Hamiltonians, FNQS can generalize to physical Hamiltonians beyond those encountered during training, offering a unified framework adaptable to various quantum systems and tasks. FNQS enable the efficient estimation of quantities that are traditionally challenging or computationally intensive to calculate using conventional methods, particularly disorder-averaged observables. Furthermore, the fidelity susceptibility can be easily obtained to uncover quantum phase transitions without prior knowledge of order parameters. These pretrained models can be efficiently fine-tuned for specific quantum systems. The architectures trained in this paper are publicly available at https://huggingface.co/nqs-models, along with examples for implementing these neural networks in NetKet.
Anatomical and histological analysis of an undescribed cervical skin fold structure in spotted turtles (Clemmys gutatta)
Replicability and generalizability of the repeated exposure effect on moral condemnation of fake news
Rapid and sustainable deep testosterone reduction predicts effective androgen deprivation therapy for metastatic hormone-sensitive prostate cancer
Deep-learning-enabled online mass spectrometry of the reaction product of a single catalyst nanoparticle
Abstract Extracting weak signals from noise is a generic challenge in experimental science. In catalysis, it manifests itself as the need to quantify chemical reactions on nanoscopic surface areas, such as single nanoparticles or even single atoms. Here, we address this challenge by combining the ability of nanofluidic reactors to focus reaction product from tiny catalyst surfaces towards online mass spectrometric analysis with the high capacity of a constrained denoising auto-encoder to discern weak signals from noise. Using CO oxidation and C 2 H 4 hydrogenation on Pd as model reactions, we demonstrate that the catalyst surface area required for online mass spectrometry can be reduced by ≈ 3 orders of magnitude compared to state of the art, down to a single nanoparticle with 0.0072 ± 0.00086 μ m 2 surface area. These results advocate deep learning to improve resolution in mass spectrometry in general and for online reaction analysis in single-particle catalysis in particular.
P-Rex2 suppresses glucose uptake into liver and skeletal muscle through different adaptor functions
Abstract P-Rex2 is a Rac guanine-nucleotide factor (Rac-GEF) that controls glucose homeostasis. This role is thought to be mediated through its adaptor function inhibiting Pten rather than through its Rac-GEF activity, but this remains to be demonstrated. To examine this question, we have investigated the roles of P-Rex2 in glucose homeostasis using Prex2 –/– and catalytically-inactive Prex2 GD mice. We show that P-Rex2 is required for insulin sensitivity but limits glucose clearance, suppressing glucose uptake into liver and skeletal muscle independently of its catalytic activity. In hepatocytes, P-Rex2 suppresses Glut2 cell surface levels, mitochondrial membrane potential and mitochondrial ATP production. We identify the orphan GPCR Gpr21 as a P-Rex2 target and propose that P-Rex2 limits hepatic glucose clearance by controlling Gpr21 trafficking. In skeletal muscle cells, P-Rex2 suppresses glucose uptake through a separate adaptor function, independently of Gpr21. Additionally, P-Rex2 suppresses insulin secretion by pancreatic islets and plasma insulin levels. Finally, P-Rex2 plays distinct Rac-GEF activity dependent and independent roles in PIP3 production in liver and skeletal muscle, respectively. Together, our study identifies complex roles of P-Rex2 in glucose homeostasis, mediated through largely GEF-activity independent mechanisms which include the GPCR Gpr21 in hepatocytes and but are not obviously linked to the regulation of Pten.
Imaging scattering resonances in low-energy inelastic ND3-H2 collisions
Association between sex and the developmental morphokinetics of in vitro derived bovine embryos
Characterizing and engineering post-translational modifications with high-throughput cell-free expression
Abstract Post-translational modifications (PTMs) are important for the stability and function of many therapeutic proteins and peptides. Current methods for studying and engineering PTMs are often limited by low-throughput experimental techniques. Here we describe a generalizable, in vitro workflow coupling cell-free gene expression (CFE) with AlphaLISA for the rapid expression and testing of PTM installing proteins. We apply our workflow to two representative classes of peptide and protein therapeutics: ribosomally synthesized and post-translationally modified peptides (RiPPs) and glycoproteins. First, we demonstrate how our workflow can be used to characterize the binding activity of RiPP recognition elements, an important first step in RiPP biosynthesis, and be integrated into a biodiscovery pipeline for computationally predicted RiPP products. Then, we adapt our workflow to study and engineer oligosaccharyltransferases (OSTs) involved in protein glycan coupling technology, leading to the identification of mutant OSTs and sites within a model vaccine carrier protein that enable high efficiency production of glycosylated proteins. We expect that our workflow will accelerate design-build-test-learn cycles for engineering PTMs.
3D mesoscopic numerical investigation on the uniaxial compressive behavior of rock-filled concrete with different ITZ and aggregate properties
Version control: how I combat the rise of generative AI in the classroom
PAM-flexible adenine base editing rescues hearing loss in a humanized MPZL2 mouse model harboring an East Asian founder mutation
The electronic structure of 1/1 ZnMgHf and its consequences for the electronic transport of the quasicrystal
Transient reactivation of small ensembles of adult-born neurons during REM sleep supports memory consolidation in mice
Abstract While memory consolidation is widely believed to require memory reactivation synchronized with theta oscillations during REM sleep, direct causal evidence linking specific neuronal ensembles to this process has been lacking. Strong theta oscillations arise in the hippocampal dentate gyrus, where a small population of principal neurons is continuously generated throughout adulthood. Although these adult-born neurons (ABNs) are known to modulate hippocampal circuits for memory, the causality between their specific information content and memory-related behavior was unknown. Here, we show that REM sleep reactivation of memory ensembles comprising as few as ~3 ABNs is necessary for fear memory consolidation. Crucially, we found that the synchronization of ABN activity with a specific theta phase is essential for associative memory consolidation. Our findings thus provide causal evidence that consolidation critically depends on both the reactivation of minimal neuronal populations and precise neuronal coordination within theta-defined temporal windows.
Engineering oriented shape optimization of GHT-Bézier developable surfaces using a meta heuristic approach with CAD/CAM applications
Scalable and lightweight deep learning for efficient high accuracy single-molecule localization microscopy
Interactive 3D segmentation for primary gross tumor volume in oropharyngeal cancer
Abstract Radiotherapy is the main treatment modality of oropharyngeal cancer (OPC), in which an accurate segmentation of primary gross tumor volume (GTVt) is essential but also challenging due to significant interobserver variability and the time consumed in manual tumor delineation. For such a challenge an interactive deep learning (DL) based approach offers the advantage of automatic high-performance segmentation with the flexibility for user correction when necessary. In this study, we investigate an interactive DL for GTVt segmentation in OPC by introducing a novel two-stage Interactive Click Refinement (2S-ICR) framework and implementing state-of-the-art algorithms. Using the 2021 HEad and neCK TumOR dataset for development and an external dataset from The University of Texas MD Anderson Cancer Center for evaluation, the 2S-ICR framework achieves a Dice similarity coefficient of 0.722 ± 0.142 without user interaction and 0.858 ± 0.050 after ten interactions, thus outperforming existing methods in both cases.
Daily briefing: This marsupial is biofluorescent — the month’s best science images
DDX41 resolves G-quadruplexes to maintain erythroid genome integrity and prevent cGAS-mediated cell death
Abstract Deleterious germline DDX41 variants constitute the most common inherited predisposition disorder linked to myeloid neoplasms (MNs), yet their role in MNs remains unclear. Here we show that DDX41 is essential for erythropoiesis but dispensable for other hematopoietic lineages. Ddx41 knockout in early erythropoiesis is embryonically lethal, while knockout in late-stage terminal erythropoiesis allows mice to survive with normal blood counts. DDX41 deficiency induces a significant upregulation of G-quadruplexes (G4), which co-distribute with DDX41 on the erythroid genome. DDX41 directly binds to and resolves G4, which is significantly compromised in MN-associated DDX41 mutants. G4 accumulation induces erythroid genome instability, ribosomal defects, and p53 upregulation. However, p53 deficiency does not rescue the embryonic death of Ddx41 hematopoietic-specific knockout mice. In parallel, genome instability also activates the cGas-Sting pathway, impairing survival, as cGas deficiency rescues the lethality of hematopoietic-specific Ddx41 knockout mice. This is supported by data from a DDX41-mutated MN patient and human iPSC-derived bone marrow organoids. Our study establishes DDX41 as a G4 resolvase, essential for erythroid genome stability and suppressing the cGAS-STING pathway.