AI News Archive: July 23, 2026 — Part 12
Sourced from 500+ daily AI sources, scored by relevance.
- Sources of Inequity and Fairness Risks inWellbeing Sensing
Passive sensing for wellbeing uses smartphones and wearables to continuously collect human behavioral data and applies ML/AI models to infer psychological states and behaviors (e.g., depression, cognitive load). These systems are increasingly adopted in high-stakes settings (e.g., hospitals, univers...
- Transparent by Design, Usable in Practice? A Formative Usability Study of a Conversational Product Advisor
Large language models can make conversational product advisors fluent but opaque. If they hide the logic behind a ranking and the evidence for a recommendation inside natural-language replies, they challenge users' ability to understand, trust, and steer the results. One response is to build transpa...
- CRAFT: Exploring Wearable Creative AI on Smart Glasses for Fiction Writing in Real-World Contexts
Creative writing increasingly integrates AI assistance, yet current tools miss in-situ moments when writers draw inspiration from real-world experiences. We envision Context-aware Reality-Fiction Transformation (CRAFT), an approach for AI glasses that translates daily experiences into fiction narrat...
- Sidewalk Moments: Are Richer Representations Always More Human-Aligned? Evidence from City-Walk Videos
We examine whether richer visual representations yield more human-aligned measures of urban engagement, using 61 first-person city-walk videos from YouTube segmented into over 50,000 ten-second clips and represented across four modalities: spatiotemporal video features, temporally averaged images (T...
- Sonic Stage: Automatically Generating Interactive Spatial Soundscapes to Facilitate Dialogue Video Comprehension for Blind Viewers
Audio description (AD) makes film and television accessible to blind and low-vision (BLV) audiences by narrating characters' actions. However, in scenes with lots of dialogue, AD often omits important actions because it is constrained not to overlap with speech. It is not yet known how to convey cha...
- Which Model Is Actually Serving You? IRIS: Budgeted Black-Box Auditing of Model Substitution and Routing Dilution in LLM Gateways
Commercial LLM gateways mediate access to hosted models, but the served backend may not match the advertised one: it may substitute a cheaper model on every request or route only a fraction $ε$ of requests to it. Prior black-box auditors often need a privileged signal (log-probabilities, token ranks...
- Toward cryptographically verifiable authorization for autonomous AI agents: A security hypothesis, preliminary formal model, and proof-of-concept implementation
Autonomous AI agents increasingly execute actions, invoke tools, and operate on protected resources with limited human oversight. Existing authentication and authorization mechanisms establish identity and delegate authority, but do not inherently provide cryptographic evidence that a concrete reque...
- Agree on the Model, Verify the Inference: GKR Protocols for HND-Based Transformer Inference
Outsourced Transformer inference exposes clients to model substitution and incomplete execution, while direct replay removes the computational benefit of delegation. We present GKR-HND, a registered-model protocol for verifying the polynomial backbone of Homomorphic--Nonhomomorphic Decomposition Tra...
- Beyond Heavy Log Curation: Perplexity-Based APT Detection via Unsupervised, Context-Augmented Language Models
Advanced Persistent Threats (APTs) remain difficult to detect because only a small fraction of events in large-scale logs are attack-related, and investigation is expensive and hard to scale. Prior machine-learning approaches can reduce analyst workload, but they often rely on heavily curated traini...
- From Resource Flow to Executable Tests: Petri-Net-Guided LLM Test Generation for Concurrent Stateful Rust APIs
Concurrent stateful library APIs expose behavior through evolving resource ownership, lifecycle states, and competing interleavings. Large language models can synthesize executable Rust tests, but their outputs often violate API preconditions, remain shallow, or reduce concurrency to accidental sequ...
- Euclid-MCP: A Model Context Protocol Server for Deterministic Logical Reasoning via Prolog
Large Language Models (LLMs) excel at natural language understanding and generation but remain unreliable for multi-step logical reasoning, especially in safety-critical or compliance-sensitive domains. Recent neuro-symbolic approaches address this gap by coupling neural models with external symboli...
- Maintenance Signals in AI-Assisted GitHub Repositories: Evidence from GenAI Adopters
Generative artificial intelligence (GenAI) can reduce code-generation effort, but it may shift work to documentation, validation, debugging, and maintenance. We study observable maintenance-cost signals among GenAI adopters on GitHub by analyzing 622 users who publicly signal adoption, 179 repositor...
- Transformer-Assisted LLM-Based Source Code Summarisation: to Enable More Secure Software Development
Neural Source Code Summarisation (NSCS) aims to generate natural language summaries of source code to improve developers' and maintainers' understanding of code. Source code summaries are vital during the maintenance phase of the Secure Software Development Lifecycle (SSDLC), as they improve maintai...
- Toward Federated Cognitive Digital Twins over the Edge-to-Cloud Continuum
Digital Twins (DTs) are increasingly adopted to monitor, analyze, and optimize Cyber-Physical Systems (CPSs) through continuous interaction between physical assets and their digital counterparts. However, current DT architectures often rely on centralized and monolithic designs, leading to scalabili...
- Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction
We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring protocol, and a cross-model leaderboard. At its core is a unified evaluation framework for constructing and running distribution-informed coding-agent ...
- From Read Speech to Spoken Digits: A Task-Specific Evaluation of Speech Privacy With Informed Attackers
Protecting speech privacy in real-life audio recordings is a growing concern. This contribution evaluates the effectiveness of three obfuscation techniques in protecting linguistic speech content, using digit recognition as a task-specific and practically motivated evaluation scenario. As a first ba...
- VibeVoice-ASR-BitNet Technical Report
We present VibeVoice-ASR-BitNet, a compressed variant of VibeVoice-ASR optimized for real-time inference on edge CPUs. We apply heterogeneous quantization tailored to the computational characteristics of each stage: the VAE acoustic tokenizer uses full-pipeline INT8 quantization (I8_S) with kernel f...
- Designed Vocalizations Dataset: Sound-Designed Human and Animal Voices for Non-human Voice Conversion
Advances in AI-based voice conversion have enabled a wide range of media applications, including films, audiobooks, and games. However, most research and public benchmarks still focus on natural human speech, leaving designed vocalizations, such as monster growls and robotic voices, underexplored, p...
- Diffusion Language Model for Recommendation
Large language model (LLM)-empowered recommender systems have emerged as a promising paradigm for generative recommendation, leveraging their strong semantic reasoning and generative capacity to model complex, diverse user preferences. However, most existing approaches rely on an autoregressive para...
- SHIFT: Self-reconstruction Harnesses Implicit Fine-grained Thinking for Retrieval
LLM-based retrievers have become a fundamental component of modern information retrieval systems. The paradigm of "rewrite-then-retriev" introduces explicit reasoning before retrieval. In addition, implicit-reasoning retrievers such as GIRCSE and LaSER improve efficiency by replacing explicit reason...
- Fast and Efficient Approximate Nearest Neighbor Search for High-Dimensional LLM Embeddings
The annual SISAP Indexing Challenge benchmarks Approximate Nearest Neighbor Search (ANNS) algorithms under rigorous constraints. This paper presents our submissions for the 2026 edition, addressing both $k$-Nearest Neighbor Graph (kNNG) construction on 1024-dimensional BGE-M3 embeddings (Task 1) and...
- Probabilistic Residual Learning for Online Recommendations
Modern recommender systems are typically based on deep learning (DL) models, where a dense encoder learns representations of users and items. As a result, these systems often suffer from the black-box nature and computational complexity of the underlying models, making it difficult to systematically...
- Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
Production AI agents' failures are less often due to an inability to reason well and more often because they cannot manage what is in their reasoning context: conversation histories, large prompts, large tool definitions, and ballooning tool outputs. Agents drown in their own accumulating history wh...
- Bridging the Structural Gap: Adapting Autoregressive Generation for Recommendation
Generative Recommendation (GR) has emerged as a new paradigm for sequential recommendation, in which a representative line of work encodes items into hierarchical semantic IDs via residual quantization and predicts the IDs token by token. However, this generative formulation still exhibits structura...
- Controllable and Content-Based Recommendations
Traditional recommendation systems rely on latent (dense) representations, making them difficult to interpret and control. We propose the Controllable and Content-Based Recommendations (CCBR) framework, which builds its recommendations from textual user profile representations. CCBR plugs into colla...
- LO-FAR: A Cost-Aware Local Filter for Sparse Feature Ranking in Industrial Ad Recommendation
Industrial ad recommendation models rely heavily on sparse, high-cardinality ID-list features that encode user histories and contextual identifiers. Each is backed by a dedicated embedding table, so these features dominate storage, training, and serving cost and must be revisited as traffic and down...
- Human rhinovirus 16 impairs macrophage cytokine secretion by disrupting NF-κB nuclear translocation and intracellular cytokine trafficking
Human rhinovirus (HRV) infections are a major cause of acute exacerbations in chronic obstructive pulmonary disease (COPD), often promoting secondary bacterial infections by dysregulating macrophage function. While we have shown that HRV16 impairs macrophage cytokine secretion, the underlying molecular mechanisms remain poorly understood. To address this, we examined the effects of HRV16 on primary human monocyte-derived macrophages, subsequently challenged with lipopolysaccharide (LPS) to mimic secondary bacterial infection. HRV16 significantly reduced IL-10 and IL-1{beta} expression at both the mRNA and protein levels. In contrast, IL-6 transcription was increased despite markedly reduced cytokine secretion. Immunofluorescence analysis revealed enhanced colocalization of IL-6 with the Golgi apparatus following HRV16 infection, consistent with intracellular retention and defective trafficking. These findings reveal that HRV16 disrupts cytokine secretion through distinct transcriptional and post-transcriptional mechanisms. To investigate the basis of transcriptional dysregulation, we performed quantitative histone post-translational modification profiling by mass spectrometry, which identified multiple HRV16-induced epigenetic alterations. Notably, a decrease in the active epigenetic mark H2AZK4Ac was observed. Chromatin immunoprecipitation demonstrated unchanged H2AZK4Ac occupancy at the IL-10 and IL-1{beta} promoters but increased enrichment at the IL-6 promoter, consistent with its selective transcriptional upregulation. HRV16 infection also induced sustained phosphorylation of NF-{kappa}B p65 that was accompanied by impaired nuclear translocation, suggesting defective activation of NF-{kappa}B-dependent transcription. Together, these results demonstrate that HRV16 inhibits cytokine secretion through disruption of NF-{kappa}B signalling and defective intracellular cytokine trafficking and identify associated alterations in the macrophage epigenetic landscape. These findings provide new mechanistic insight into how HRV impairs macrophage inflammatory responses and may compromise antibacterial immunity during COPD exacerbations.
- Genetic Diversity and Antifungal Susceptibility among Ecological Niche Populations of Aspergillus flavus in Yaounde, Cameroon
Aspergillus flavus is a ubiquitous fungus commonly found in a variety of ecological niches including soil, crops, the air, and humans. In humans, it is the second leading cause of invasive aspergillosis (IA) and is linked to other illnesses through contamination of agricultural products with aflatoxins. As such, A. flavus threatens food safety, economic wellbeing, and human health, particularly in less developed regions and countries such as Cameroon. To mitigate these effects, farmers and clinicians rely on triazoles to reduce aflatoxin contamination and treat IA. However, a consequence of increasing triazole use is the emergence and spread of triazole resistance in both agriculture and clinics. To identify the prevalence of triazole resistance and the potential relationships among triazole resistant and susceptible strains, this study investigated antifungal susceptibility to both clinical and agricultural triazoles and analyzed genetic variation among various ecological niche populations of A. flavus in Cameroon. Strain genotypes were obtained through analysis of six polymorphic microsatellite markers. Our analyses revealed that 30.3% (17/56) of the strains were resistant to at least one of the four triazoles, with increased minimum inhibitory concentrations found among crop-isolated strains. A Permutational Multivariate Analysis of Variance suggested limited ecological niche-based clustering of genotypes, consistent with frequent gene flow and dispersal among ecological niches. A multilocus linkage disequilibrium analysis revealed evidence of non-random recombination in A. flavus. Overall, this study elucidates the interplay between ecological pressures, antifungal resistance, and genetic differentiation, and invites alternative methods to control aflatoxin contamination in foods without the use of agricultural fungicides.
- A genetic screen unveils the polymerization cycle of the bacterial actin-like MreB
MreB, a bacterial actin homologue that organizes cell wall synthesis in most rod-shaped bacteria, assembles into membrane-associated filaments through mechanisms that have remained elusive despite decades of study. In particular, how ATP binding, ATP hydrolysis and membrane association are coordinated during the MreB polymerization cycle has remained unknown. Here, we combine genetics with systematic biochemical characterizations of purified MreB variants to decipher the key molecular steps governing the MreB assembly cycle. A genetic screen in Bacillus subtilis identified essential residues controlling filament assembly and membrane association. Systematic biochemical analyses of purified homologous MreB variants from Geobacillus stearothermophilus demonstrated that monomer-monomer interactions are required for both ATP hydrolysis and high-affinity membrane binding, whereas ATP binding, but not ATP hydrolysis, is sufficient to promote polymerization. Conversely, ATP hydrolysis destabilizes MreB polymers, promoting filament turnover. Together, these results support a model in which ATP-binding induces an early nucleation step that increases membrane affinity, membrane association promotes filament assembly, and subsequent ATP hydrolysis completes the cycle by driving polymer disassembly and turnover. These findings provide a mechanistic basis for understanding how MreB polymerization dynamics may regulate the spatial and temporal organization of bacterial cell wall growth.
- RBM6-mediated alternative splicing facilitates the adaptation of Eurasian avian-like H1N1 swine influenza virus
The Eurasian avian-like (EA) H1N1 swine influenza virus (SIV), derived from avian influenza viruses (AIV), poses a serious threat to public health due to its capacity for cross-species transmission and pandemic emergence. The molecular determinants underlying its replication advantage over AIV remain poorly defined. Here, we identify RNA-binding motif protein 6 (RBM6) as a novel host factor that differentially regulates the replication of EA H1N1 SIV and AIV. Mechanistically, RBM6 binds to the critical M901 site of the viral M segment, thereby modulating RNA splicing. Substitution of M901C with M901T markedly reduced RBM6 binding, impaired M segment splicing, and attenuated viral replication both in vitro and in vivo. Conservation analysis revealed that M901T is common in avian strains, whereas M901C is predominantly maintained in swine strains, underscoring M901C as a determinant of swine adaptation. Complementation experiments further demonstrated that swine RBM6, but not avian RBM6, restored EA H1N1 SIV replication. Taken together, our findings uncover a previously unrecognized role of RBM6 in shaping influenza virus replication and highlight the RBM6-M901C axis may serve as potential targets for controlling influenza virus adaptation and interspecies transmission.
- Plasmodium blood-stage induces trained immunity in hepatocytes
Plasmodium parasites, the causative agents of malaria, are transmitted at high levels in endemic areas and sequential infections are common. Using a mouse model of infection we discovered liver burden was suppressed in blood-stage experienced, compared to naive, animals, independently of adaptive responses and inflammation. We observed greater chromatin accessibility of a subset of interferon stimulated genes in blood-stage experienced animals, and a rapid increase in transcription of these genes upon sporozoite challenge. Ex vivo stimulation of hepatocytes from blood-stage experienced mice also led to a rapid and elevated response upon treatment with an unrelated antigen, ctDNA. Taken together, these data are consistent with a model in which hepatocytes are reprogramed by blood stage Plasmodium infection to exhibit trained memory akin to what has been described in innate immune cells. The consequences of training of hepatocytes could be wide-reaching and might alter responses to diverse pathogens and other stimuli.
- Microbial odours activate protective immune response in Caenorhabditis elegans via specific olfactory neurons
Caenorhabditis elegans, like other animals, relies on attractive odours for foraging and on aversive odours for avoidance of pathogens. An odour produced by pathogenic bacterium Pseudomonoas aeruginosa induces immune response as well as avoidance in C. elegans via AWB olfactory neurons. We asked whether AWB neurons provide broader immunity to a wide range of pathogens. We activated AWB neurons using three chemically distinct bacterial odours and found that activation indeed induces protective immunity. Conversely, silencing of AWB neurons early during infection enhances susceptibility to infections. Mechanistic investigation revealed that the activation of AWB using odours upregulates detoxification pathways and other protective pathways in non-neuronal tissues. Specifically, odour exposure activates UDP-glucuronosyltransferase UGT-18 in the intestine that protects worms from the phenazine toxin of Pseudomonas aeruginosa and promotes broader immunity to Gram-negative bacteria (P. aeruginosa and Salmonella enterica), Gram-positive bacteria, (Enterococcus faecalis and Staphylococcus aureus), and yeast (Cryptococcus neoformans). Altogether, our findings highlights microbial odours as non-canonical molecular patterns for activating immune responses.
- SARS-CoV-2 nucleocapsid protein engages with viral RNA and ERGIC lipids to drive viral core assembly
Severe acute respiratory syndrome coronavirus 2 assembles at the ER Golgi intermediate compartment (ERGIC), yet the molecular basis of nucleocapsid (N) protein interactions with host membranes remains unclear. Using in vitro reconstituted lipid membranes and viral RNA N complexes, we show that full-length N binds phosphatidylinositol (PI) and phosphatidylserine (PS) containing membranes and induces lipid clustering, an effect amplified by viral RNA and ERGIC like membrane lipid composition. The isolated N-terminal domain lacks this activity, whereas the C terminal domain retains membrane-associated multimerization. Although, lipid bilayers promote co-clustering of N and PI lipids, facilitating ribonucleoprotein (RNP) assembly, even on simple membranes. Importantly, ERGIC-mimicking membranes enhances this co clustering further stabilizing RNPs of dimensions matching the viral core. In cells, viral RNA enhances N clustering without altering particle production. These findings reveal cooperative interactions between N, viral RNA, and ERGIC lipids as key drivers of lipid-dependent viral core formation, providing a mechanistic framework for the early steps of viral assembly.
- The Metabarcoding Analysis Pipeline (MAP): Simple, accurate, and flexible metabarcoding
Current metabarcoding pipelines are inflexible with respect to study design and are poorly suited to long-read sequence data. To address these limitations, we developed MAP, the Metabarcoding Analysis Pipeline, which is a sequence-to-answer workflow supporting the analysis of amplicons from highly multiplexed and replicated study designs. Although MAP can analyze amplicons of any length from any genetic marker, it includes several features tailored to long-read COI metabarcoding. MAP installs from a Docker container and requires only sequence data, a parameters file, and a reference library. It produces intuitive reports, enabling users to evaluate their data immediately after analysis. We validate MAP by showing that it generates biodiversity estimates that correspond closely to a ground-truth dataset of single-specimen DNA barcode data and by demonstrating that it outperforms alternative platforms for COI metabarcoding. MAP is free, open-source, and available from: https://github.com/cbg-innov/MAP.
- Epigenetic Regulation of Stable SARS-CoV-2 RBD-sfGFP Expression in Primary Human Splenic Fibroblasts
Background: Recombinant expression of the SARS-CoV-2 receptor-binding domain (RBD) is essential for vaccine development, serological diagnostics, and mechanistic studies. Primary human fibroblasts offer physiologically relevant protein folding and post-translational modification, yet their short lifespan limits scalable production. We used an immortalized human splenic fibroblast cell line to stably express RBD-sfGFP for longitudinal characterization and downstream studies. Methods: Immortalized human primary splenic fibroblasts were transfected by electroporation with a plasmid encoding SARS-CoV-2 RBD fused to superfolder GFP (sfGFP), with a neomycin resistance cassette (neoR) for G418 selection. Four independent G418-resistant cultures (n=4), designated HPSF-IM-RBD-BHSKPU T1-T4, were established from distinct selection flasks. Based on previous screenings, two cultures (T1, T3) were monitored for 98 days (14 passages, P1-P14); two cultures (T2, T4) were monitored for 42 days (6 passages, P1-P6). RBD-sfGFP expression was assessed by fluorescence microscopy at 7-day intervals. For each timepoint, 2 fields were imaged and analyzed for relative fluorescence intensity (normalized to global maximum = 100%) and mean fluorescence intensity (MFI, normalized to global maximum = 100%). Coefficient of variation (CV), linear regression, and Pearson correlation were calculated. Results: All four cultures exhibited robust GFP fluorescence, confirming stable transgene retention. Expression ranking: T1 (93.1% +/- 3.6%) > T3 (89.2% +/- 3.4%) > T2 (84.2% +/- 3.2%) > T4 (79.7% +/- 3.9%). Long-term cultures T1 and T3 retained ~100% of Day 7 signal at Day 98 (T1: 100.7%; T3: 100.0%). Expression exhibited passage-dependent oscillation rather than progressive silencing. CV increased over time in T1 (1.5% -> 8.5%), indicating growing intercellular heterogeneity. A strong positive correlation between fluorescence and MFI (Pearson r = 0.823, p = 7.44 x 10^-11) suggested coherent population-level regulation. Conclusions: HPSF-IM-RBD-BHSKPU cells stably retain RBD-sfGFP expression for over 3 months, validating their utility as a recombinant protein production platform. However, oscillatory dynamics and increasing heterogeneity are consistent with position-effect variegation at distinct integration loci. Consequently, early passages (P1-P4) are optimal for applications requiring maximal uniformity. Ultimately, these cells provide a practical tool for RBD production and a valuable model for studying epigenetic regulation of transgene expression in human primary fibroblast backgrounds.
- Antibody evasion and receptor binding of SARS-CoV-2 variants PQ.16.1.1 and RK.1
The recent global expansion of the SARS-CoV-2 variant NB.1.8.1 has driven the emergence of sublineages PQ.16.1.1 and RK.1, which independently acquired the D420N mutation in their receptor-binding domains and and now dominate the Asia-Pacific region. Evaluations utilizing surface plasmon resonance and pseudovirus assays demonstrate that these sublineages exhibit significantly reduced human ACE2 receptor engagement compared to their parental strain. However, this functional cost is offset by a marked ability to evade humoral immunity, specifically demonstrating profound resistance to Class 1 neutralizing monoclonal antibodies and convalescent plasma from Wuhan-Hu-1-primed individuals. This convergent evolution exemplifies a classical viral trade-off, sacrificing receptor binding efficiency to escape population-level immune pressure. Consequently, these findings suggest these variants will soon spread globally and emphasize the critical need for ongoing surveillance to monitor D420N-carrying lineages.
- Mediterranean dispersal imposed bottlenecks in Neolithic sheep and goats
The Neolithic period in western Eurasia involved major dispersal events, including the translocation of livestock into Europe. These domestic animal translocations likely featured founder effects, fundamentally reshaping their genetic landscape. One well-known example is the Mediterranean dispersal, hypothesised to involve rapid maritime translocation with small founding herds, yet the genetic consequences for the livestock themselves remain poorly understood. To investigate this we generated 12 palaeogenomes of sheep and goat from La Draga and Cova de l'Avellaner, and coanalyzed these with other Neolithic Mediterranean assemblages, including 4 additional novel genomes. We observe a significantly lower genetic diversity and higher runs of homozygosity (ROH) in Iberian sheep and goat populations compared to other ancient European and southwest Asian populations. We quantify the magnitude of the bottleneck in goats at ~205 Ne, highlighting the severity of the founder event. We further identify species-specific demographic trajectories: goats experienced a severe and prolonged bottleneck, likely driven by regional isolation within the western Mediterranean, intensive exploitation strategies, and small initial founding herds, while sheep were likely more frequently exchanged across the region. Lastly, identity-by-descent (IBD) indicates connectivity along the Mediterranean coast compatible with a cabotage model. This suggests that seaways facilitated rapid initial dispersal but not sustained livestock transfer from southwest Asia, shaping the genetic foundation of Iberian livestock herds.
- Sex-biased transcriptional rewiring of neuronal circuits associated with eprinomectin resistance in Haemonchus contortus
Anthelmintic resistance poses a critical threat to livestock health, with clinical failures of eprinomectin and other macrocyclic lactones documented in the gastrointestinal parasite Haemonchus contortus worldwide. We performed whole-genome Pool-seq of larvae and RNA-seq on adult male and female worms from five phenotypically characterised isolates - two eprinomectin-susceptible and three resistant - from dairy sheep in Southwest France. We identified a conserved macrocyclic lactone resistance locus on chromosome 5 and a putatively eprinomectin-specific locus on chromosome 4, alongside sex-specific transcriptomic responses linking genomic selection signatures to transcriptional phenotype. Predicted functions of selected and differentially expressed genes support reorganisation of neuronal circuits in drug responsiveness, including the kynurenine pathway producing neuroactive metabolites, further supported by quinolinic acid potentiating eprinomectin efficacy in larval assays. Despite geographic proximity, resistance appears to have evolved independently across farms. Together, these findings reveal a complex genetic architecture underlying eprinomectin resistance with implications for targeted diagnostics and resistance management.
- Non-invasive tracking of ripple-like activity during human sleep using MEG
Hippocampal ripples are considered a key mechanism of sleep-dependent memory consolidation. In humans, however, their direct investigation has relied on invasive recordings from patients with epilepsy, because EEG provides limited access to fast activity from deep medial temporal structures. Here, we tested whether source-resolved magnetoencephalography can reveal physiologically meaningful ripple-like activity during NREM sleep in healthy humans. These events showed hallmarks expected of sleep ripples: they were most prominent in medial temporal regions and embedded within slow oscillation (SO)-spindle dynamics, occurring preferentially during SO up-states and spindle centers. To test mnemonic relevance, we combined sleep MEG with targeted memory reactivation, in which learning cues were replayed during sleep. TMR improved memory performance, and later remembered cues preferentially recruited medial temporal ripple-like events alongside cortical spindles. Ripple-like events also coincided with enhanced item-specific reactivation during successful TMR. Together, these findings establish MEG as a non-invasive window onto hippocampo-cortical memory dynamics during sleep.
- Movement Directions Aligned in Joint Space Are Not Aligned in Muscle Space
Learned movements are thought to be represented in both extrinsic and intrinsic coordinate systems. Intrinsic representations have traditionally been characterized using joint-based coordinates, although the relationship between joint movements and muscle activation depends strongly on limb configuration. Consequently, movement directions aligned in joint space may not be aligned in muscle space, but the implications of this mismatch for motor learning have remained largely unexplored. We addressed this question by combining electromyographic (EMG) analysis with a visuomotor adaptation experiment. In Experiment 1, participants performed planar reaching movements in two workspaces separated by a 45 degree shoulder rotation while EMG activity was recorded from nine upper-limb muscles. Muscle-pattern similarity analysis revealed that movement directions aligned in joint space were not always aligned in muscle space and that the degree of misalignment varied systematically across movement directions. Based on these results, we predicted that visuomotor adaptation to clockwise (CW) and counterclockwise (CCW) rotations would produce different patterns of motor generalization, contrary to the prediction of conventional joint-space accounts. Experiment 2 confirmed this prediction, revealing a systematic shift between the CW and CCW generalization patterns that was consistent with the muscle-space prediction. These findings suggest that intrinsic representations of learned movements are not fully captured by joint-based coordinates alone and that muscle-based coordinates contribute to motor learning and its generalization. Together, these findings highlight the importance of considering underlying biomechanics when interpreting motor representations using generalization paradigms.
- Endometrial transcriptomic and ciliation analysis after scratch shows no signature linked to live birth following IVF
Objective: To determine whether endometrial scratch induces differences in transcriptomic profiles or epithelial cell ciliation in the endometrium at the window of implantation that associate with live birth following IVF. Design: Secondary analysis of 50 matched endometrial biopsies collected within a randomised controlled trial evaluating the clinical effectiveness of endometrial scratch before first time IVF. Setting: Endometrial biopsy samples were obtained from women attending the Jessop Wing of Sheffield Teaching Hospitals. Population or Sample: Women undergoing first time IVF who received an endometrial scratch in the preceding cycle at the window of implantation (6-10 days after LH surge). Methods: Endometrial biopsies were molecularly dated using transcriptomic menstrual cycle staging algorithms. Bulk RNA sequencing was analysed using DESeq2 with FDR correction, and principal component analysis (PCA) assessed clustering patterns. Epithelial cell ciliation was quantified using immunohistochemistry and an automated Python based image analysis pipeline. Main Outcome Measures: Differential endometrial gene expression between women with and without live birth after IVF; percentage coverage of ciliated epithelial cells in luminal and glandular regions. Results: Transcriptomic dating confirmed no differences in menstrual cycle timing between live birth and no live birth groups. No significant differential gene expression was detected (log2FC >2, FDR <0.05), and PCA showed no clustering by pregnancy outcome. Ciliation coverage did not differ between outcome groups or between glandular and luminal surfaces. Conclusions: When implantation timing is precisely defined, endometrial scratch does not produce detectable transcriptomic changes or alterations in epithelial ciliation that distinguish women who achieve live birth after IVF. Funding: Wellbeing of Women RG2147; Wessex Medical Research; Rosetrees Trust. Keywords: Endometrial scratch, IVF, transcriptomics, epithelial ciliation, window of implantation, bulk RNA sequencing, endometrial receptivity.
- Symmetry-breaking flow bifurcation as an under-recognised haemodynamic factor in valve-associated thrombosis
Venous thrombosis commonly develops in the vicinity of venous valve pockets, where disturbed haemodynamics, xreduced washout, and endothelial dysfunction promote thrombus initiation. While previous studies have focused on conventional flow descriptors such as velocity, shear stress, recirculation, and residence time, the influence of valve mechanics on flow organisation remains poorly understood. Here, we combine a biomimetic vein-on-a-chip platform with computational fluid dynamics, fluid-structure interaction simulations, Ghost Particle Velocimetry (GPV), whole-blood flow measurements, and particle transport experiments to investigate the interplay between valve biomechanics and venous haemodynamics. Movable venous valve leaflets fabricated by in situ photopolymerisation of poly(ethylene glycol) diacrylate (PEGDA) enabled independent control of leaflet stiffness under physiologically relevant steady and pulsatile flow conditions. Previous experiments demonstrated that leaflet flexibility governs thrombus localisation, with symmetric leaflet stiffness promoting clot formation at the valve tips, while asymmetric stiffness shifts thrombus formation towards the valve sinus. In addition, we identify a previously undescribed Reynolds-number-dependent symmetry-breaking transition in post-valve flow. Above a critical flow condition, an initially symmetric jet spontaneously develops into a stable asymmetric flow pattern. This behaviour was consistently reproduced experimentally using GPV and confirmed by Fluid-structure simulations. Valve compliance delayed the onset of the transition by increasing the effective leaflet opening, whereas valves with a small geometric offset promoted earlier asymmetry through higher local flow velocities within the valve gap. The resulting asymmetric flow generated persistent lateral bias in the transport of red blood cell-sized particles, suggesting enhanced platelet accumulation and prolonged residence within one valve sinus. These findings demonstrate that venous valve mechanics regulate not only local flow fields but also particle transport relevant to thrombosis. The discovery of a stable symmetry-breaking flow state provides a new haemodynamic mechanism linking valve stiffness, asymmetric particle transport, and the preferential localisation of thrombus formation, offering new insight into the mechanobiology of deep vein thrombosis.
- Nanopore Electrometry Resolves Peptide Charge Patterns beyond Ionic-Current Blockade
Localized measurements of electric fields offer a promising route to expand the information content of nanopore-based single-molecule sensing beyond conventional ionic-current blockade. Here, using all-atom molecular dynamics simulations with virtual electric-field probes placed around a graphene nanopore, we show that the local electric-field captures the presence, and distribution of charged amino acids as the peptides translocate through the pore. These field signatures create reproducible peptide-specific fingerprints across independent translocation events and enable substantially improved discrimination between peptides compared with ionic-current traces obtained under the same simulation conditions. Our results suggest that localized nanopore electrometry can provide a complementary, information-rich readout of peptide charge order that is largely inaccessible to conventional current-blockade-based measurements. This study establishes a simulation-guided framework for integrating nanoscale electrometry with nanopore platforms for future peptide and protein analysis.
- PhytoFam: A Nextflow Pipeline for Genome-Wide Analysis of Plant Gene Families
Genome-wide identification of plant gene families is essential for functional and evolutionary studies but often requires the use of multiple independent tools for homolog detection, domain validation, orthology assignment, and phylogenetic analysis. This fragmented approach involves extensive manual scripting, complicates reproducibility and parameter tracking, and may require additional steps to remove redundant protein isoforms. To address these challenges, we developed PhytoFam, a Nextflow-based workflow that automates gene family identification from proteome input through phylogenetic reconstruction. The pipeline integrates HMMER for candidate sequence identification, isoform-aware deduplication, InterProScan for domain confirmation, BLAST reciprocal best hit (RBH) analysis for orthology assignment, MUSCLE for multiple sequence alignment with optional outgroup incorporation, TrimAl for alignment trimming, and IQ-TREE3 for phylogenetic reconstruction. PhytoFam is portable across local workstations and high-performance computing environments and supports deployment through Conda, Docker, and Singularity. We validated the workflow using the Morus alba MADS-box gene family, where the complete analysis finished in 1 h 10 min (9 CPU h). IQ-TREE3 accounted for most of the execution time, whereas InterProScan showed the highest memory requirement with a peak resident set size of 4.5 GB. PhytoFam provides a reproducible, automated, and scalable solution for plant gene family identification and phylogenetic analysis. The pipeline is freely available at https://github.com/sanamparajuli/PhytoFam.
- A highly penetrant LMNA R541C variant associated with dilated cardiomyopathy leads to dysregulation in metabolism and proliferation pathways in stem cell-derived cardiomyocytes
Background LMNA codes a widely expressed nuclear cytoskeletal protein (lamin A/C) with multiple important functions. Pathogenic LMNA genetic variation may lead to autosomal dominant cardiomyopathy, though the severity and rate of progression can vary with the specific nucleotide change and location. Prior studies showed that induced pluripotent stem cells (iPSC)-derived cardiomyocytes (iCMs) with LMNA R541C exhibited reduced LMNA protein abundance, increased sarcomere disorganization, and abnormal electrophysiology. Methods We investigated the LMNA-R541C variant that exhibits a highly penetrant and severe clinical cardiomyopathy phenotype using transcriptomic analysis of iCMs. Patient-derived iPSCs with CRISPR-corrected (clustered regularly interspersed short palindromic repeats) isogenic control cells and CRISPR knock-in LMNA-R541C heterozygous iPSCs were generated for isogenic controlled experiments. Results In differential gene expression analyses we observed that LMNAR541C/WT iPSC-derived cardiomyocytes had consistent perturbations in 123 genes across CRISPR-corrected and knock-in experiments compared to controls. Pathway analysis identified that the G2M checkpoint and oxidative phosphorylation processes were consistently dysregulated and confirm these findings in previously published iPSC and murine models. Discussion These results implicate perturbed gene expression and pathways that may contribute to the severe phenotypes in LMNA-R541C. Informatic analysis of pathways suggests several drug classes including multiple cardiac glycosides as potential targeted therapeutic candidates to be explored.
- Programmed chromosome elimination correlates with the overexpression of cohesin and additional B chromosome-encoded genes in Aegilops speltoides
Programmed chromosome elimination is a highly controlled developmental process in which specific chromosomes are selectively lost from defined cell types during development. Despite its broad occurrence across plants and animals, the molecular mechanisms driving the tissue-specific elimination of chromosomes remain largely unresolved. Here, we exploit the root-specific elimination of supernumerary B chromosomes in Aegilops speltoides as a tractable model to identify the genetic basis of programmed chromosome loss. A high-quality, chromosome-scale genome assembly was generated, assigning 398 Mb of sequence to the Ae. speltoides B chromosome. Transcriptome profiling across seven tissue types representing chromosome elimination-active, elimination-negative, and B chromosome nondisjunction conditions identified 3,262 genes consistently upregulated in elimination-associated tissues, including 1,035 B genes. Stepwise subtraction of genes expressed in post-elimination and B chromosome-retaining reference tissues, followed by intersection with genes expressed during B nondisjunction in anthers, identified a candidate gene set enriched for chromosome segregation functions. From this set, we prioritized SYN2-B, a B chromosome-encoded cohesin -kleisin subunit whose Arabidopsis thaliana ortholog AtSYN2 induces chromosome bridges and micronucleus formation when misregulated. CENH3-B, an -type centromeric histone variant identified through GO enrichment analysis of B genes expressed in elimination-associated tissues, was shown to be incorporated in centromeres of both A and B chromosomes by transient gene expression assays using protoplasts and 3D structured illumination microscopy. These findings support a model in which B chromosome-encoded perturbations of cohesin activity and centromere composition contribute to selective B chromosome nondisjunction and their elimination in root tissues. Moreover, the SYN2-B promoter is enriched for ethylene response factor-binding sites compared to its A-encoded paralog, suggesting that ethylene is implicated in the root identity pathway driving root-specific B chromosome elimination.
- Adversarial Validation Reveals Diagnostic Workflow Leakage in PCOS Machine Learning Models
Background: Machine-learning models for polycystic ovary syndrome (PCOS) and other conditions frequently report near-perfect diagnostic performance, but retrospective datasets assembled from routine clinical practice can encode diagnostic-group membership in how data were acquired rather than in disease biology, and this acquisition-related information can be indistinguishable from genuine clinical signal under conventional validation. Objective: To determine, using a real-world PCOS cohort as a case study, whether high classification performance reflected clinically meaningful information or artifacts of data provenance, schema structure, and measurement-acquisition workflow, and to develop a generalizable audit framework for detecting such artifacts in retrospective medical machine learning. Methods: We analyzed 1,331 retrospective records (1,286 PCOS, 45 controls) from a single endocrine-gynecology database. A layered acquisition-bias framework compared classification performance using (i) raw and harmonized missingness patterns alone, (ii) measured values with and without explicit missingness indicators, and (iii) ascertainment-balanced feature sets with and without age. Logistic regression and random forest were evaluated using repeated stratified cross-validation, bootstrap resampling, label-permutation testing, and calibration analysis, and the framework was validated against a semi-synthetic experiment with known ground truth. Results: Diagnostic status was perfectly predicted (ROC-AUC = 1.000) from missingness patterns alone, before any clinical value was examined, and this persisted after semantic harmonization of duplicated source columns. Performance declined progressively as acquisition-sensitive information was removed, from near-ceiling in raw and harmonized value models to a mean ROC-AUC of approximately 0.80-0.82 in the most restrictive ascertainment-balanced, age-excluded representation. The semi-synthetic experiment reproduced this pattern under known data-generating conditions, confirming that harmonization removes schema-fragmentation artifacts but not workflow-driven acquisition bias. Conclusions: Apparent diagnostic performance in this cohort was substantially attributable to diagnostic workflow and data-acquisition structure rather than to a stable, transportable biological signal. The layered audit framework generalizes beyond PCOS and offers a practical tool for detecting acquisition-related leakage in retrospective clinical machine-learning studies.
- Machine learning-based neuroimaging for prediction of deep brain stimulation outcomes in movement disorders: Systematic review and meta-analysis
Abstract Background: Deep Brain Stimulation (DBS) surgery is a treatment of choice for movement disorders, and utilizes an implanted electrical pulse generator that administers electrical stimulation to designated brain regions responsible for motor control. The preoperative identification of effective predictive factors is of utmost importance for appropriate patient selection. In this study, we evaluate the potential of machine learning-based neuroimaging for predicting DBS outcomes. (PROSPERO Registration: CRD420261279318) Method: Following the PRISMA statement, eligible studies were selected through searching three databases (PubMed, Scopus, Web of Science) on November 6, 2025. Methodological quality was assessed using the PROBAST+AI tool. Random-effects models pooled discrimination performance (AUC). Heterogeneity was investigated using meta-regressions for age and gender alongside subgroup analysis by type of algorithm. Publication bias was assessed using Egger regression test. Results: Twenty studies were included in the analysis. Most investigations focused on PD, STN-DBS, and postoperative motor improvement, while a smaller number assessed neuropsychiatric outcomes. Overall, the pooled discrimination for models predicting motor outcomes showed an AUC of 0.86, and the pooled models for delirium showed an AUC of 0.87. Regarding the risk of bias assessment, seven studies were classified as low risk, while thirteen were identified as high risk. Conclusion: Machine learning-based neuroimaging shows promising potential for preoperative prediction of DBS outcomes. However, the current literature is characterized by a persistent gap between encouraging discrimination and reliable clinical readiness. The main weakness of the field lies in analytical rigor and generalizability. These models should currently only be considered as promising research tools.
- Risk stratification for the rapid pain progression phenotype in knee osteoarthritis using interpretable multimodal machine learning: Development in the Osteoarthritis Initiative and external evaluation in the Prospective Cohort of Osteoarthritis from A Coruna
Objective. To develop an interpretable multimodal machine-learning model for risk stratification of the rapid pain progression phenotype in knee osteoarthritis and to evaluate its performance in the independent PROCOAC cohort. Methods. An elastic-net logistic regression model was trained using Osteoarthritis Initiative (OAI) data. Rapid pain progression was defined over overlapping 24-month windows using normalized WOMAC pain. Harmonized clinical, genetic and proteomic candidates were evaluated, with feature selection by permutation importance. The frozen algorithm was tested in an OAI hold-out set and externally evaluated in PROCOAC. Logistic recalibration corrected prevalence shifts. Clinical utility was assessed by decision curve analysis. Results. OAI comprised 2,934 individuals and 14,488 instances. Feature pruning reduced 159 candidates to a 19-variable clinical-genetic signature driven by Kellgren-Lawrence grade, localized knee pain, BMI and two genetic variants (rs73631790, rs9912678); no proteomic variable was retained. External testing in PROCOAC (582 individuals, 1609 instances) showed ROC-AUC 0.744 (95% CI 0.714 to 0.772) and PR-AUC 0.519. Following recalibration, the sensitive screening threshold yielded NPV 0.875 (95% CI 0.849 to 0.898) and sensitivity 0.804 (95% CI 0.760 to 0.844), whereas the high-specificity threshold achieved PPV 0.610 (95% CI 0.523 to 0.692) and specificity 0.941 (95% CI 0.924 to 0.954). Decision curve analysis showed positive net benefit at both thresholds, supporting a three-tier risk stratification framework. Conclusions. This externally evaluated model identified patients at risk of rapid pain progression using an MRI-free clinical-genetic signature. Recalibrated thresholds may support risk-adapted monitoring, advanced imaging prioritization and trial enrichment.
- Cross-country generalizability of foundation models for cervical cancer screenings on H&E whole slide images
Accurate grading of cervical biopsies on Hematoxylin and Eosin (H&E) stained whole slide images (WSIs) is essential for distinguishing high grade lesions from low grade changes, yet this process is subject to considerable inter-observer variability. In this study, we evaluate a foundation model-based multiple instance learning (MIL) pipeline for binary high-grade squamous intraepithelial lesion (HSIL) detection on H&E stained WSIs. We benchmark our Athena foundation model against four state-of-the-art pathology foundation models: H-optimus-0, Hibou-L, Midnight-12k and Virchow, across datasets from five different countries: Portugal, Cambodia, Germany, Poland and Scotland. Athena achieved the highest mean area under the curve (AUC) (0.931) with the lowest cross-country variability (STD = 0.022). Furthermore, we compared the model's diagnostic performance to that of trained pathologists on a dataset with p16-confirmed ground truth. Our model improved sensitivity from 84% to 95% while maintaining comparable specificity (85% vs. 84%). Failure analysis revealed that the model's errors were concentrated at the diagnostic boundary between low-grade and high-grade lesions, whereas pathologists' errors spanned a broader range of misclassifications. These findings show the potential of foundation models for cervical cancer screenings worldwide.