❌

Reading view

Mitophagy-related gene signatures predict prognosis and therapeutic response in hepatocellular carcinoma

Biochem Biophys Res Commun. 2026 Sep 30;838:154646. doi: 10.1016/j.bbrc.2026.154646. Online ahead of print.

ABSTRACT

Mitophagy, a selective form of autophagy, has been implicated in tumor progression and therapeutic resistance; however, its prognostic significance in hepatocellular carcinoma (HCC) remains unclear. In this study, we comprehensively evaluated the role of mitophagy-related genes in HCC using multi-omics data. Gene expression profiles were obtained from the TCGA-LIHC and GSE14520 cohorts, and mitophagy-related genes were retrieved from the GeneCards database. Twenty differentially expressed mitophagy-related genes with prognostic value (pDEMGs) were identified, and consensus clustering stratified HCC patients into two clusters with significantly different survival outcomes (P = 0.001). A mitophagy enrichment score (MIES) was then calculated using single-sample gene set enrichment analysis (ssGSEA). Elevated MIES was associated with poorer overall survival (HR = 2.17, P = 0.005), metabolic activation, immune suppression, and differential drug sensitivity. Single-cell analysis of the GSE140228 dataset revealed heterogeneous MIES activity across cell populations, with relatively higher enrichment observed in proliferating T cells and dendritic cells. A six-gene prognostic signature (ACTR6, GAPDH, ATIC, ANP32E, CCT6A, and BSG) was developed using LASSO-Cox regression, which effectively stratified patients into high- and low-risk groups with distinct overall survival outcomes (1-, 3-, and 5-year AUCs: 0.780, 0.682, and 0.690, respectively). The risk score was correlated with immune infiltration patterns, mutational landscape, and chemotherapy response. qPCR validation further confirmed the upregulation of ACTR6, CCT6A, ATIC, and BSG in HCC cells. Collectively, these findings establish a mitophagy-related scoring system that reflects immune and genomic characteristics, as well as a six-gene signature with independent prognostic value, highlighting the potential clinical relevance of mitophagy in HCC.

PMID:42828884 | DOI:10.1016/j.bbrc.2026.154646

  •  

Author Correction: Intermittent hypobaric pressure induces selective senescent cell death and alleviates age-related osteoporosis

Nature Biomedical Engineering, Published online: 28 September 2026; doi:10.1038/s41551-026-01815-3

Author Correction: Intermittent hypobaric pressure induces selective senescent cell death and alleviates age-related osteoporosis
  •  

Mechanism of Action of Hedyotis diffusa Extract in a Rat Model of Acute Lung Injury Based on Transcriptomic Analysis

Biology (Basel). 2026 Sep 4;15(17):1549. doi: 10.3390/biology15171549.

ABSTRACT

OBJECTIVE: This study established a rat model of lipopolysaccharide (LPS)-induced acute lung injury (ALI) to evaluate pathological damage, collagen deposition, inflammatory cytokine levels, and key gene/protein expression following Hedyotis diffusa water extract (HDWE) intervention. Combined with ultra-high-performance liquid chromatography-quadrupole Orbitrap high-resolution mass spectrometry (UHPLC-Q-Orbitrap HRMS), transcriptomic analysis, and molecular simulation, this study identified the bioactive components of HDWE, evaluated their potential interactions with ALI-related targets, and explored the multi-omics-based protective mechanisms of HDWE.

METHODS: Thirty-six Sprague-Dawley (SD) rats were randomly divided into six groups: Control group, ALI group, DXMS group, HDWE-L group (100 mg/kg), HDWE-M group (200 mg/kg), and HDWE-H group (300 mg/kg). Hematoxylin and eosin (H&E) and Masson's trichrome staining were used to evaluate lung pathological changes and collagen deposition. Enzyme-linked immunosorbent assay (ELISA) was used to measure serum tumor necrosis factor-α TNF-α interleukin-1β IL-1β, erleukin-6 (IL-6), and interleukin-10 (IL-10) levels. Transcriptomic analysis identified differentially expressed genes (DEGs), followed by Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), receiver operating characteristic (ROC), and immune infiltration analyses. Quantitative real-time polymerase chain reaction (qRT-PCR) detected the mRNA expression levels of SPHK1, RELA, and NFKBIA. Immunohistochemistry evaluated the expression of eight hub targets, including endothelin-1 (EDN1), sphingosine kinase 1 (SPHK1), intercellular adhesion molecule 1 (ICAM1), interleukin-17 (IL-17), prostaglandin-endoperoxide synthase 2 (PTGS2/COX-2), NF-κB p65 (encoded by RELA), WT1-associated protein (WTAP), and myeloperoxidase (MPO). UHPLC-Q-Orbitrap HRMS characterized HDWE constituents. Molecular docking analysis was performed between 22 compounds and eight hub targets, followed by 100 ns molecular dynamics simulations and molecular mechanics-Poisson-Boltzmann surface area (MM/PBSA) binding free energy calculations for five core targets. Compared with the control group, the ALI group showed increased levels of TNF-α (86%), IL-1β (107%), and IL-6 (66%), accompanied by a 43% reduction in IL-10 and a 300% increase in lung collagen deposition. All HDWE doses alleviated inflammatory responses, with medium-dose HDWE showing the most pronounced effects. Specifically, medium-dose HDWE increased IL-10 levels by 52% and reduced IL-6, TNF-α, and IL-1β levels by 18%, 22%, and 11%, respectively. Transcriptomic analysis identified 2512 DEGs between the control group and ALI groups, 832 exclusive DEGs between the ALI group and HDWE-M groups, and 876 overlapping DEGs enriched in TNF, IL-17, and NF-κB signaling pathways. The eight-hub-gene diagnostic model achieved an area under the curve (AUC) of 0.969. RELA, SPHK1, and four other hub genes showed positive correlations with Th1, Th17, and neutrophil infiltration. In the ALI group, SPHK1, RELA, and NFKBIA mRNA expression levels were 1.30-, 0.96-, and 0.71-fold of those in the control group, respectively. Compared with the ALI group, high-dose HDWE treatment and low-dose HDWE treatment reduced SPHK1 expression to 0.62- and 0.57-fold, respectively, and increased NFKBIA expression to 1.68- and 1.58-fold, respectively. High-dose HDWE treatment reduced RELA expression to 0.43-fold. The expression levels of inflammation-related proteins were increased in the ALI group and were reduced after HDWE treatment. Twenty-two HDWE components were identified, 16 of which met the docking criteria. Asperulosidic acid exhibited favorable predicted binding affinities with all eight targets, with calculated binding free energies of -14.74, -14.92, -17.58, -23.04, and -16.10 kcal/mol for MPO, IL-17, NF-κB p65, PTGS2/COX-2, and SPHK1, respectively.

CONCLUSIONS: This study provides systematic in vivo pharmacodynamic and in silico component-target evidence regarding the protective effects of HDWE against LPS-induced ALI. HDWE treatment increased NFKBIA expression and reduced SPHK1, RELA, and multiple inflammatory protein levels, suggesting that HDWE may regulate the IL-17/NF-κB-associated inflammatory network, although direct causal relationships require further validation. Asperulosidic acid may represent a key bioactive component with broad target-binding potential. This study was limited by the use of an LPS-induced rat ALI model without gene knockout or target inhibitor validation; therefore, further functional experiments are required to confirm the proposed regulatory mechanisms.

PMID:42737981 | PMC:PMC13564518 | DOI:10.3390/biology15171549

  •  

When Does AI Augment Work? A Workflow-Level Framework for Human-Agent Collaboration

arXiv:2609.12482v1 Announce Type: new Abstract: We aim to characterise the value of artificial intelligence in the workplace. Current studies largely measure this value in terms of the current automation capabilities and public adoption of AI. However, such metrics ignore the greater impacts of human--agent collaboration in transforming the nature of work. To account for this, we must expand the scope of our analysis beyond atomised tasks of today, and instead focus on how AI can augment entire workflows of the future. To ground this analysis, we establish a precise definition of AI augmentation comprising six conditions, spanning durable net value, meaningful human control, accountability and recovery, and long-term human development through learning, career pathways, and job purpose. We elaborate on these conditions and apply the framework in a case study of AI-mediated social surveys. We conclude by outlining how organisations, researchers, and government leaders can use this framework to make sense of the future of work.
  •  

Sensing Intelligence as a Trainable Metamaterial Property

arXiv:2605.23967v1 Announce Type: new Abstract: In biological systems, sensing is not performed by the brain alone: the body deforms, vibrates, and filters external stimuli before they are transduced into neural signals. In engineered systems, this processing burden is placed largely on electronics and computation, while the mechanical body is usually designed only for strength and stability. Here, we present sensing intelligence as a trainable property of the body. We show that the geometry of a metamaterial can be optimized to reshape external stimuli into internal signals that are easier for a neural network to interpret. Rather than hand-designing this physical preprocessing, we let the neural network train its own body for sensing by backpropagating the sensing loss to the body's design parameters through differentiable simulation. Across numerical and experimental sensing scenarios, the optimized body improves sensing accuracy by up to fivefold or reduces the number of required electronic sensors by nearly an order of magnitude.
  •  

Deubiquitinase USP35 regulates MDM4 degradation to promote endothelial ferroptosis and renal injury progression

Cell Death Discovery, Published online: 25 May 2026; doi:10.1038/s41420-026-03128-5

Deubiquitinase USP35 regulates MDM4 degradation to promote endothelial ferroptosis and renal injury progression
  •  

IC3-Evolve: Proof-/Witness-Gated Offline LLM-Driven Heuristic Evolution for IC3 Hardware Model Checking

arXiv:2604.03232v1 Announce Type: new Abstract: IC3, also known as property-directed reachability (PDR), is a commonly-used algorithm for hardware safety model checking. It checks if a state transition system complies with a given safety property. IC3 either returns UNSAFE (indicating property violation) with a counterexample trace, or SAFE with a checkable inductive invariant as the proof to safety. In practice, the performance of IC3 is dominated by a large web of interacting heuristics and implementation choices, making manual tuning costly, brittle, and hard to reproduce. This paper presents IC3-Evolve, an automated offline code-evolution framework that utilizes an LLM to propose small, slot-restricted and auditable patches to an IC3 implementation. Crucially, every candidate patch is admitted only through proof- /witness-gated validation: SAFE runs must emit a certificate that is independently checked, and UNSAFE runs must emit a replayable counterexample trace, preventing unsound edits from being deployed. Since the LLM is used only offline, the deployed artifact is a standalone evolved checker with zero ML/LLM inference overhead and no runtime model dependency. We evolve on the public hardware model checking competition (HWMCC) benchmark and evaluate the generalizability on unseen public and industrial model checking benchmarks, showing that IC3-Evolve can reliably discover practical heuristic improvements under strict correctness gates.
  •  

Do Emotions in Prompts Matter? Effects of Emotional Framing on Large Language Models

arXiv:2604.02236v1 Announce Type: new Abstract: Emotional tone is pervasive in human communication, yet its influence on large language model (LLM) behaviour remains unclear. Here, we examine how first-person emotional framing in user-side queries affect LLM performance across six benchmark domains, including mathematical reasoning, medical question answering, reading comprehension, commonsense reasoning and social inference. Across models and tasks, static emotional prefixes usually produce only small changes in accuracy, suggesting that affective phrasing is typically a mild perturbation rather than a reliable general-purpose intervention. This stability is not uniform: effects are more variable in socially grounded tasks, where emotional context more plausibly interacts with interpersonal reasoning. Additional analyses show that stronger emotional wording induces only modest extra change, and that human-written prefixes reproduce the same qualitative pattern as LLM-generated ones. We then introduce EmotionRL, an adaptive emotional prompting framework that selects emotional framing adaptively for each query. Although no single emotion is consistently beneficial, adaptive selection yields more reliable gains than fixed emotional prompting. Together, these findings show that emotional tone is neither a dominant driver of LLM performance nor irrelevant noise, but a weak and input-dependent signal that can be exploited through adaptive control.
  •  

Editing strigolactone hormone receptor for robust antiviral silencing in rice

Precise genome editing of the rice strigolactone receptor DWARF14 confers robust, transgene-free antiviral resistance by blocking viral suppression of endogenous RNA silencing, offering a promising strategy for durable disease protection without a yield penalty.
  •  

AgentSLR: Automating Systematic Literature Reviews in Epidemiology with Agentic AI

arXiv:2603.22327v1 Announce Type: cross Abstract: Systematic literature reviews are essential for synthesizing scientific evidence but are costly, difficult to scale and time-intensive, creating bottlenecks for evidence-based policy. We study whether large language models can automate the complete systematic review workflow, from article retrieval, article screening, data extraction to report synthesis. Applied to epidemiological reviews of nine WHO-designated priority pathogens and validated against expert-curated ground truth, our open-source agentic pipeline (AgentSLR) achieves performance comparable to human researchers while reducing review time from approximately 7 weeks to 20 hours (a 58x speed-up). Our comparison of five frontier models reveals that performance on SLR is driven less by model size or inference cost than by each model's distinctive capabilities. Through human-in-the-loop validation, we identify key failure modes. Our results demonstrate that agentic AI can substantially accelerate scientific evidence synthesis in specialised domains.
  •  

SRAM-Based Compute-in-Memory Accelerator for Linear-decay Spiking Neural Networks

arXiv:2603.12739v1 Announce Type: cross Abstract: Spiking Neural Networks (SNNs) have emerged as a biologically inspired alternative to conventional deep networks, offering event-driven and energy-efficient computation. However, their throughput remains constrained by the serial update of neuron membrane states. While many hardware accelerators and Compute-in-Memory (CIM) architectures efficiently parallelize the synaptic operation (W x I) achieving O(1) complexity for matrix-vector multiplication, the subsequent state update step still requires O(N) time to refresh all neuron membrane potentials. This mismatch makes state update the dominant latency and energy bottleneck in SNN inference. To address this challenge, we propose an SRAM-based CIM for SNN with Linear Decay Leaky Integrate-and-Fire (LD-LIF) Neuron that co-optimizes algorithm and hardware. At the algorithmic level, we replace the conventional exponential membrane decay with a linear decay approximation, converting costly multiplications into simple additions while accuracy drops only around 1%. At the architectural level, we introduce an in-memory parallel update scheme that performs in-place decay directly within the SRAM array, eliminating the need for global sequential updates. Evaluated on benchmark SNN workloads, the proposed method achieves a 1.1 x to 16.7 x reduction of SOP energy consumption, while providing 15.9 x to 69 x more energy efficiency, with negligible accuracy loss relative to original decay models. This work highlights that beyond accelerating the (W x I) computation, optimizing state-update dynamics within CIM architectures is essential for scalable, low-power, and real-time neuromorphic processing.
  •  

HKDC1-Mediated Polyamine Rewiring Drives Lenvatinib Resistance and Immune Escape in Hepatocellular Carcinoma

Clin Mol Hepatol. 2026 Mar 11. doi: 10.3350/cmh.2025.1269. Online ahead of print.

ABSTRACT

BACKGROUND/AIMS: Lenvatinib resistance and immune exclusion limit outcomes in HCC. We hypothesized that metabolic rewiring orchestrates resistance to lenvatinib and PD-1 blockade.

METHODS: We established LS/LR HCC models and employed multi-omics (proteomics/RNA-seq), ChIP, luciferase, and RIP assays to map HKDC1 regulation. Tumor immunity was profiled by scRNA-seq, mIHC, and flow cytometry. SPD + lenvatinib efficacy was tested in cell lines, patient-derived organoids/xenografts. Tested therapy effect in an immunocompetent hydrodynamic HCC model with hepatocyte-specific Hkdc1 deletion; and analyzed a postoperative cohort (n = 40) treated with lenvatinib + PD-1.

RESULTS: HKDC1, upregulated in LR HCC, was transcriptionally activated by USF1 and promoted SMS-mediated polyamine rewiring. This impaired CD8⁺ T-cell metabolism, reversible by HKDC1 knockdown or spermidine (SPD). SPD synergized with lenvatinib, triggering autophagy and suppressing tumor growth in vitro and in vivo. High HKDC1 predicted poor response and survival in patients receiving lenvatinib + aPD-1.

CONCLUSIONS: A USF1/HKDC1/SMS axis couples polyamine metabolism to immune dysfunction and lenvatinib resistance. HKDC1 is a predictive biomarker and therapeutic node and support polyamine-axis modulation to sensitize HCC to lenvatinib plus PD-1 therapy.

PMID:41812646 | DOI:10.3350/cmh.2025.1269

  •  

Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents

arXiv:2509.26354v2 Announce Type: replace Abstract: Advances in Large Language Models (LLMs) have enabled a new class of self-evolving agents that autonomously improve through interaction with the environment, demonstrating strong capabilities. However, self-evolution also introduces novel risks overlooked by current safety research. In this work, we study the case where an agent's self-evolution deviates in unintended ways, leading to undesirable or even harmful outcomes. We refer to this as Misevolution. To provide a systematic investigation, we evaluate misevolution along four key evolutionary pathways: model, memory, tool, and workflow. Our empirical findings reveal that misevolution is a widespread risk, affecting agents built even on top-tier LLMs (e.g., Gemini-2.5-Pro). Different emergent risks are observed in the self-evolutionary process, such as the degradation of safety alignment after memory accumulation, or the unintended introduction of vulnerabilities in tool creation and reuse. To our knowledge, this is the first study to systematically conceptualize misevolution and provide empirical evidence of its occurrence, highlighting an urgent need for new safety paradigms for self-evolving agents. Finally, we discuss potential mitigation strategies to inspire further research on building safer and more trustworthy self-evolving agents. Our code and data are available at https://github.com/ShaoShuai0605/Misevolution . Warning: this paper includes examples that may be offensive or harmful in nature.
  •  

Ready2Unlearn: A Learning-Time Approach for Preparing Models with Future Unlearning Readiness

arXiv:2505.10845v2 Announce Type: replace-cross Abstract: Machine unlearning is the process of removing the imprint left by specific data samples during the training of a machine learning model. AI developers, including those building personalized technologies, employ machine unlearning for various purposes such as privacy protection, security, and to address ethical concerns. This paper introduces Ready2Unlearn, a learning-time optimization approach designed to facilitate future unlearning processes. Unlike the majority of existing unlearning efforts that focus on designing unlearning algorithms, which are typically implemented reactively when an unlearning request is made during the model deployment phase, Ready2Unlearn shifts the focus to the training phase, adopting a "forward-looking" perspective. Building upon well-established meta-learning principles, Ready2Unlearn proactively trains machine learning models with unlearning readiness, such that they are well prepared and can handle future unlearning requests in a more efficient and principled manner. Ready2Unlearn is model-agnostic and compatible with any gradient ascent-based machine unlearning algorithms. We evaluate the method on both language and vision tasks under various unlearning settings, including class-wise unlearning and random data unlearning. Experimental results show that by incorporating such preparedness at training time, Ready2Unlearn produces an unlearning-ready model state, which offers several key advantages when future unlearning is requested. We hope this study inspires future research on proactive strategies for equipping machine learning models with built-in unlearning readiness, particularly in modern information systems that rely heavily on user data for recommendation, search, and personalized services, where privacy risks and data deletion demands are increasingly prevalent.
  •  

HumanLM: Simulating Users with State Alignment Beats Response Imitation

arXiv:2603.03303v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used to simulate how specific users respond to a given context, enabling more user-centric applications that rely on user feedback. However, existing user simulators mostly imitate surface-level patterns and language styles, which fail to reflect the underlying states of real users (e.g., beliefs and emotions). To address these limitations, we propose a novel training framework, HumanLM, which builds user simulators that accurately reflect real users. Our key insight is that, in addition to generating responses, the model should generate natural-language latent states that align with ground-truth responses through reinforcement learning. These latent states correspond to a set of psychologically grounded state dimensions that drive how real users respond. HumanLM further synthesizes these aligned latent states into responses that accurately represent real users. For extensive evaluation, we develop Humanual, a comprehensive benchmark for simulating real users based on public data. Humanual consists of six large-scale datasets with 26k users and 216k responses in total, spanning diverse tasks such as generating user responses to daily life issues, political blogs, and chat sessions with LLM assistants. Across datasets, HumanLM significantly outperforms alternative approaches, achieving an average relative improvement of 16.3% in alignment scores from an LLM judge. In a real-time simulation study with 111 participants, HumanLM achieves the highest similarity to real user responses and competitive human-likeness scores.
  •  

Kaleido: Open-Sourced Multi-Subject Reference Video Generation Model

arXiv:2510.18573v2 Announce Type: replace-cross Abstract: We present Kaleido, a subject-to-video~(S2V) generation framework, which aims to synthesize subject-consistent videos conditioned on multiple reference images of target subjects. Despite recent progress in S2V generation models, existing approaches remain inadequate at maintaining multi-subject consistency and at handling background disentanglement, often resulting in lower reference fidelity and semantic drift under multi-image conditioning. These shortcomings can be attributed to several factors. Primarily, the training dataset suffers from a lack of diversity and high-quality samples, as well as cross-paired data, i.e., paired samples whose components originate from different instances. In addition, the current mechanism for integrating multiple reference images is suboptimal, potentially resulting in the confusion of multiple subjects. To overcome these limitations, we propose a dedicated data construction pipeline, incorporating low-quality sample filtering and diverse data synthesis, to produce consistency-preserving training data. Moreover, we introduce Reference Rotary Positional Encoding (R-RoPE) to process reference images, enabling stable and precise multi-image integration. Extensive experiments across numerous benchmarks demonstrate that Kaleido significantly outperforms previous methods in consistency, fidelity, and generalization, marking an advance in S2V generation.
  •  

Contextualized Privacy Defense for LLM Agents

arXiv:2603.02983v1 Announce Type: cross Abstract: LLM agents increasingly act on users' personal information, yet existing privacy defenses remain limited in both design and adaptability. Most prior approaches rely on static or passive defenses, such as prompting and guarding. These paradigms are insufficient for supporting contextual, proactive privacy decisions in multi-step agent execution. We propose Contextualized Defense Instructing (CDI), a new privacy defense paradigm in which an instructor model generates step-specific, context-aware privacy guidance during execution, proactively shaping actions rather than merely constraining or vetoing them. Crucially, CDI is paired with an experience-driven optimization framework that trains the instructor via reinforcement learning (RL), where we convert failure trajectories with privacy violations into learning environments. We formalize baseline defenses and CDI as distinct intervention points in a canonical agent loop, and compare their privacy-helpfulness trade-offs within a unified simulation framework. Results show that our CDI consistently achieves a better balance between privacy preservation (94.2%) and helpfulness (80.6%) than baselines, with superior robustness to adversarial conditions and generalization.
  •  

Comparing AI Agents to Cybersecurity Professionals in Real-World Penetration Testing

arXiv:2512.09882v2 Announce Type: replace Abstract: We present the first comprehensive evaluation of AI agents against human cybersecurity professionals in a live enterprise environment. We evaluate ten cybersecurity professionals alongside six existing AI agents and ARTEMIS, our new agent scaffold, on a large university network consisting of ~8,000 hosts across 12 subnets. ARTEMIS is a multi-agent framework featuring dynamic prompt generation, arbitrary sub-agents, and automatic vulnerability triaging. In our comparative study, ARTEMIS placed second overall, discovering 9 valid vulnerabilities with an 82% valid submission rate and outperforming 9 of 10 human participants. While existing scaffolds such as Codex and CyAgent underperformed relative to most human participants, ARTEMIS demonstrated technical sophistication and submission quality comparable to the strongest participants. We observe that AI agents offer advantages in systematic enumeration, parallel exploitation, and cost -- certain ARTEMIS variants cost $18/hour versus $60/hour for professional penetration testers. We also identify key capability gaps: AI agents exhibit higher false-positive rates and struggle with GUI-based tasks.
  •  
❌