Headlines

A new study published today in Stem Cell Reports demonstrates how a human stem cell-derived model of the intestine can be used to identify potential therapies for inflammatory bowel disease (IBD), highlighting glycyrrhizin as a promising candidate for reducing intestinal inflammation and cell death.

Fewer than 10 people worldwide have eradicated the virus with stem cells. But this case was special—no one knew his brother's cells carried a protective mutation until transplant day. The post Norwegian Man Cured of HIV by His Brother’s Stem Cells appeared first on SingularityHub.

In a groundbreaking investigation poised to reshape nutritional recommendations for older adults, a new study has revealed a compelling relationship between adherence to the Mediterranean diet and the incidence of dyspeptic symptoms in elderly populations. Conducted by Otay Lule, N., Lule, K.O., and Ozturk, Z.A., the research delves deep into how dietary patterns can influence […]

In the relentless battle against breast cancer, the challenge of preventing tumor recurrence after surgical intervention has remained a formidable obstacle. Traditionally, post-surgical chemotherapy and tissue reconstruction have often been compromised by the limitations of scaffold-based materials that fail to integrate seamlessly with native tissues or degrade at a pace compatible with tissue regeneration. However, […]

Restricting calorie intake in species such as mice, rhesus monkeys, and fruit flies has been shown to extend their lifespans. In some cases, these animals not only live longer, but are also free of disease. But when pushed too far, calorie restriction can have negative impacts. Mice that undergo a 40% reduction in calorie intake, for example, are more susceptible to infections, less likely to reproduce, and experience stunted growth.

AI is a great tool for small, well-defined tasks, but maintain a healthy skepticism and avoid falling down a rabbit hole.

The National Institute on Aging's Interventions Testing Program (ITP) is the full stop at the end of many a debate over the merits of development of one substance or another as a hoped for treatment to modestly slow aging. The ITP uses a very large number of mice and considerable rigor to assess effects on life span. The program typically focuses on small molecules and supplements that have prior evidence for anti-aging effects, and usually those with a long history in the literature. Given the number of compounds that show no effect on life span in the hands of the ITP, this initiative serves as a reminder that any one study in a hundred mice that demonstrates modest slowing of aging does not in fact […]

In the quest to unravel the mysteries of aging, blood has emerged not just as a biomarker reflecting the passage of time but as an active player capable of modulating the aging process itself. A groundbreaking study recently published in Experimental & Molecular Medicine sheds light on the intricate mechanisms by which blood influences aging […]

Researchers have developed a modified version of the CRISPR gene-editing tool that in early laboratory experiments suggests it may have the potential to “silence” the extra...

A transformative wave is sweeping through modern medicine, driven by advances in gene and RNA-based therapies that promise not merely to manage diseases but to correct them at their genetic roots. Despite the immense therapeutic potential, a persistent challenge has hampered progress: safely and precisely delivering these molecular treatments to their intended cellular destinations, particularly […]

In a groundbreaking development that could redefine the future of organ transplantation, researchers have unveiled the results of a pioneering clinical trial investigating a novel approach to managing immune response after living-donor liver transplantation. The innovative strategy centers around the infusion of donor-derived regulatory dendritic cells (DCreg) combined with the early withdrawal of conventional immunosuppressive […]

Two previously unknown stem cell types appear to play a central role in children's height growth, according to a study from the University of Gothenburg. The study also shows that growth hormone can act directly on these cells.

Both physicians and the general public want AI in health care to be more accurate than humans. At the same time, trust in the technology is only moderate. That is the finding of a study from the University of Gothenburg.

New Treatment Lets 3 Transplant Patients Halt Anti-Rejection Drugs  The New York TimesFirst-of-its-kind Pitt study allows some liver transplant recipients to stop antirejection drugs  Pittsburgh Post-GazetteCell therapy primed liver transplant patients to avoid organ rejection, small study shows  statnews.comFirst-in-human trial primes immune system to accept donor livers  Medical XpressFirst-Ever Human Trial Trains Immune System to Accept Donor Livers  Bioengineer.org

New Treatment Lets 3 Transplant Patients Halt Anti-Rejection Drugs  The New York TimesFirst-of-its-kind Pitt study allows some liver transplant recipients to stop antirejection drugs  Pittsburgh Post-GazetteCell therapy primed liver transplant patients to avoid organ rejection, small study shows  statnews.comFirst-in-human trial primes immune system to accept donor livers  Medical XpressFirst-Ever Human Trial Trains Immune System to Accept Donor Livers  Bioengineer.org

The landscape of cardiovascular medicine is on the brink of a transformative evolution, propelled by the advent of soft robotic technologies. Traditional devices and simulators employed in cardiovascular disease management have long struggled with inherent limitations — their rigidity, lack of biomimetic fidelity, and the invasive nature of surgical procedures often hinder clinical outcomes. In […]

Heart transplant surgeon Raymond Lee, MD, explains what patients can expect after a heart transplant—including how long the heart will last. The average lifespan of a transplanted heart is about 10 years, though outcomes vary from patient to patient.

Ученые из Рокфеллерского университета представили способ превращения иммунной системы в место производства лекарственных белков. Их эксперименты продемонстрировали, что генетическое редактирование позволяет перепрограммировать иммунные клетки так, чтобы они сами производили терапевтические молекулы прямо в организме, включая сложные антитела, которые трудно синтезировать традиционными методами, https://medicalxpress.com/news/2026-04-immune-therapeutic-proteins.html Medical Express. Новая технология может изменить лечение огромного числа заболеваний от рака до

One of the major projects within the study of comparative biology is the attempt to understand why adult individuals of some species can fully regenerate lost tissues following injury, while mammals such as our own species cannot. A variety of modest inroads into identifying potentially important differences in cellular biochemistry and activity have been made, such as work focused on senescent cells and macrophages, but it remains an unsolved challenge. Researchers here present more data to add to that already under consideration, focused on the role of oxygen sensing in the initial response to injury. It is unclear as to whether it can lead to dramatic improvements in mammalian regeneration, but the work suggests that regeneration could be improved via manipulation of oxygen sensing in […]

Researchers at the University of Pittsburgh gave transplant recipients certain immune cells from their organ donors. It didn’t always work.

Researchers at the University of Pittsburgh gave transplant recipients certain immune cells from their organ donors. It didn’t always work.

What do sheep, regeneration of Maine farmland, and indigenous knowledge have to do with one another? Quite a bit, when it comes to Melissa Schneider, shepherd and landowner of Casa Portal. Along with herding and sustainable land management, Melissa is one of many women farmers who are staking their place. The post On a Maine Farm, Regeneration Starts With the Sheep appeared first on Modern Farmer.

Named after the crystallographer who helped reveal the structure of DNA, GPT-Rosalind is OpenAI’s first domain-specific model series, fine-tuned for biochemistry, genomics, and protein engineering. Access is restricted to a trusted-access programme for vetted enterprise customers including Amgen, Moderna, and Thermo Fisher Scientific. OpenAI has launched GPT-Rosalind, a frontier reasoning model built specifically for life […] This story continues at The Next Web

arXiv:2604.15269v1 Announce Type: cross Abstract: The impossibility of simultaneously cloning non-orthogonal states lies at the foundations of quantum theory. Even when allowing for approximation errors, cloning an arbitrary unknown pure state requires as many initial copies as needed to fully learn the state. Rather than arbitrary unknown states, modern quantum learning theory often considers structured classes of states and exploits such structure to develop learning algorithms that outperform general-state tomography. This raises the question: How do the sample complexities of learning and cloning relate for such structured classes? We answer this question for an important class of states. Namely, for $n$-qubit stabilizer states, we show that the optimal sample complexity of cloning is $\Theta(n)$. Thus, also for this structured class of states, cloning is as hard as learning. To prove these results, we use representation-theoretic tools in the recently proposed Abelian State

arXiv:2604.14484v1 Announce Type: cross Abstract: Behavior cloning (BC) policies on position-controlled robots inherit the closed-loop response of the underlying PD controller, yet the effect of controller gains on BC failure lacks a nonasymptotic theory. We show that independent sub-Gaussian action errors propagate through the gain-dependent closed-loop dynamics to yield sub-Gaussian position errors whose proxy matrix $X_\infty(K)$ governs the failure tail. The probability of horizon-$T$ task failure factorizes into a gain-dependent amplification index $\Gamma_T(K)$ and the validation loss plus a generalization slack, so training loss alone cannot predict closed-loop performance. Under shape-preserving upper-bound structural assumptions the proxy admits the scalar bound $X_\infty(K)\preceq\Psi(K)\bar X$ with $\Psi(K)$ decomposed into label difficulty, injection strength, and contraction, ranking the four canonical regimes with compliant-overdamped (CO) tightest, stiff-underdamped

arXiv:2604.15269v1 Announce Type: cross Abstract: The impossibility of simultaneously cloning non-orthogonal states lies at the foundations of quantum theory. Even when allowing for approximation errors, cloning an arbitrary unknown pure state requires as many initial copies as needed to fully learn the state. Rather than arbitrary unknown states, modern quantum learning theory often considers structured classes of states and exploits such structure to develop learning algorithms that outperform general-state tomography. This raises the question: How do the sample complexities of learning and cloning relate for such structured classes? We answer this question for an important class of states. Namely, for $n$-qubit stabilizer states, we show that the optimal sample complexity of cloning is $\Theta(n)$. Thus, also for this structured class of states, cloning is as hard as learning. To prove these results, we use representation-theoretic tools in the recently proposed Abelian State

arXiv:2604.14202v1 Announce Type: new Abstract: Electroencephalography (EEG) has become one of the key modalities underpinning brain-computer interfaces (BCIs) due to its high temporal resolution, rapid responsiveness, non-invasiveness, low cost, and portability. However, EEG signals are substantially inferior to intracranial EEG (iEEG) in signal-to-noise ratio and local spatial resolution, whereas iEEG suffers from extremely limited clinical accessibility owing to its invasive nature, hindering widespread application. To address this challenge, this study proposes a unified data-and prior knowledge-driven framework for EEG-iEEG representational enhancement. Guided by the principle that "geometric structure dictates function", the framework maps static cortical anatomy onto dynamic constraints governing neural signal propagation and integrates general-purpose neural representations extracted by a pre-trained large EEG model to explicitly model signal transmission through the brain.

arXiv:2604.15269v1 Announce Type: cross Abstract: The impossibility of simultaneously cloning non-orthogonal states lies at the foundations of quantum theory. Even when allowing for approximation errors, cloning an arbitrary unknown pure state requires as many initial copies as needed to fully learn the state. Rather than arbitrary unknown states, modern quantum learning theory often considers structured classes of states and exploits such structure to develop learning algorithms that outperform general-state tomography. This raises the question: How do the sample complexities of learning and cloning relate for such structured classes? We answer this question for an important class of states. Namely, for $n$-qubit stabilizer states, we show that the optimal sample complexity of cloning is $\Theta(n)$. Thus, also for this structured class of states, cloning is as hard as learning. To prove these results, we use representation-theoretic tools in the recently proposed Abelian State

arXiv:2604.14202v1 Announce Type: cross Abstract: Electroencephalography (EEG) has become one of the key modalities underpinning brain-computer interfaces (BCIs) due to its high temporal resolution, rapid responsiveness, non-invasiveness, low cost, and portability. However, EEG signals are substantially inferior to intracranial EEG (iEEG) in signal-to-noise ratio and local spatial resolution, whereas iEEG suffers from extremely limited clinical accessibility owing to its invasive nature, hindering widespread application. To address this challenge, this study proposes a unified data-and prior knowledge-driven framework for EEG-iEEG representational enhancement. Guided by the principle that "geometric structure dictates function", the framework maps static cortical anatomy onto dynamic constraints governing neural signal propagation and integrates general-purpose neural representations extracted by a pre-trained large EEG model to explicitly model signal transmission through the brain.

arXiv:2604.14585v1 Announce Type: new Abstract: Prompt optimization in compound AI systems is statistically indistinguishable from a coin flip: across 72 optimization runs on Claude Haiku (6 methods $\times$ 4 tasks $\times$ 3 repeats), 49% score below zero-shot; on Amazon Nova Lite, the failure rate is even higher. Yet on one task, all six methods improve over zero-shot by up to $+6.8$ points. What distinguishes success from failure? We investigate with 18,000 grid evaluations and 144 optimization runs, testing two assumptions behind end-to-end optimization tools like TextGrad and DSPy: (A) individual prompts are worth optimizing, and (B) agent prompts interact, requiring joint optimization. Interaction effects are never significant ($p > 0.52$, all $F

arXiv:2604.14484v1 Announce Type: new Abstract: Behavior cloning (BC) policies on position-controlled robots inherit the closed-loop response of the underlying PD controller, yet the effect of controller gains on BC failure lacks a nonasymptotic theory. We show that independent sub-Gaussian action errors propagate through the gain-dependent closed-loop dynamics to yield sub-Gaussian position errors whose proxy matrix $X_\infty(K)$ governs the failure tail. The probability of horizon-$T$ task failure factorizes into a gain-dependent amplification index $\Gamma_T(K)$ and the validation loss plus a generalization slack, so training loss alone cannot predict closed-loop performance. Under shape-preserving upper-bound structural assumptions the proxy admits the scalar bound $X_\infty(K)\preceq\Psi(K)\bar X$ with $\Psi(K)$ decomposed into label difficulty, injection strength, and contraction, ranking the four canonical regimes with compliant-overdamped (CO) tightest, stiff-underdamped (SU)

Patagonia shareholder Holdfast Collective will assist Rodale Institute's Regenerate Ventura project for farmers hoping to transition. The post Patagonia & Rodale Institute to transition an entire California county to regenerative agriculture appeared first on AgFunderNews.

A new gene‑editing strategy programs hematopoietic stem cells to produce therapeutic antibodies and other proteins. The approach generates long‑lasting immunity in mice and could point to single‑vaccine treatments for diverse diseases in the future. The post Stem Cell Editing Programs the Immune System to Make Own Therapeutic Proteins appeared first on GEN - Genetic Engineering and Biotechnology News.

This Perspective reviews how AI is moving drug discovery from long, costly experimental pipelines toward earlier clinical translation, with an AI-designed TNIK inhibitor trial serving as an important proof-of-feasibility reference point. It argues that precision oncology could benefit from AI-driven integration of multi-omics, federated learning, and adaptive trial design, but only if validation, fairness, interpretability, and regulatory alignment improve.

Researchers audited five popular public-facing AI chatbots across 250 health prompts and found that 49.6% of responses were problematic, with especially weak performance on open-ended questions and in nutrition and athletic performance. The study also found poor citation quality, difficult readability, and very few refusals to answer, raising concerns about the reliability of chatbots for everyday health information.

The immune system acts as a critical sentinel of organismal aging, integrating the sensing of physiological states with the execution of defense and clearance functions. Immunosenescence not only reflects systemic functional decline but also serves as a central driver of multiple age-related diseases. However, owing to the high heterogeneity and complexity of the immune system, accurately quantifying its aging status and identifying actionable intervention targets have been major scientific challenges in the field.

Gallbladder cancer (GBC) is a rare form of cancer with no signs or symptoms in the early stages. In the U.S., approximately 2,000 people die annually from this condition, with only 20% diagnosed at an early stage. Surgery remains the most effective treatment. Although minimally invasive approaches—laparoscopic and robotic—are increasingly used in gastrointestinal oncology, their use in GBC is limited and comparing robotic surgery to laparoscopic and conventional surgery approaches remains limited and controversial.

Megan Morrone / Axios: OpenAI launches GPT-Rosalind, an AI model for life sciences research, including drug discovery, as a research preview for customers such as Moderna and Amgen  —  OpenAI announced a new series of AI models built to help life sciences researchers work faster.

In a groundbreaking advancement that may redefine the future landscape of HIV treatment, researchers have reported the achievement of long-term remission of HIV-1 in a patient who underwent an allogeneic hematopoietic stem cell transplant (HSCT) from a sibling donor harboring the CCR5Δ32/Δ32 mutation. This striking clinical success, detailed in a recent publication in Nature Microbiology, […]

Gallbladder Cancer and the Rise of Minimally Invasive Surgery: A Transformative Approach Gallbladder cancer (GBC), although rare, poses a formidable challenge in oncology due to its silent early progression. In the United States alone, it claims approximately 2,000 lives annually, with a mere 20% of cases detected early enough to allow for curative interventions. Historically, […]

Agrifood corporates now need tools to help them assess the financial impact of transitioning supply chains to regenerative agriculture practices. The post Behind Klim’s new tool to turn regenerative agriculture into a financial instrument appeared first on AgFunderNews.

Cells have evolved responses to stress that enhance the chance of survival. Many of these responses converge of increased activity of maintenance processes, more recycling of materials, less protein synthesis, and a number of other common mechanisms. Researchers have found that mild stress of near any sort imposed upon a living organism will provoke a net gain in cell function and resilience, which in turn acts to modestly slow progression of the complicated cascade of accumulating damage and dysfunction that we call aging. The bounds of the possible are illustrated by the response to the nutrient stress, induced by fasting or calorie restriction. Short-lived mammalian species such as mice can live as much as 40% longer in response to a restricted but still sufficient nutrient […]

Researchers developed an improved method for creating insulin-producing cells from human stem cells, which effectively regulated blood sugar levels in laboratory tests and reversed diabetes in a mouse model. The post Advances in Stem Cell‑Derived Insulin‑Producing Cells for Type 1 Diabetes appeared first on GEN - Genetic Engineering and Biotechnology News.

Stem cells proliferate after injury to repair damaged tissue, and chronic injury can promote cancer. However, the injury-activated signals and regulatory mechanisms, and their relationship to cancer, are poorly understood. Here, we identified insulin-...

Long-term in vivo production of therapeutic proteins and development of vaccines that elicit protective levels of broadly neutralizing antibodies (bNAbs) against major pathogens face challenges. In this study, we report on an alternative gene editing ...

The UK government’s £500m Sovereign AI fund announces first cohort of startups backed to boost economic growth and national security

Ventura County will host a two-field series about regenerative ag this May, bringing together growers, ag researchers and sustainability leaders to explore the future of the region’s specialty crops. The series will be presented in partnership with the Ecological Farming Association, and will allow attendees to examine practical approaches to regenerative farming across avocados, citrus … The post Ventura County Farmers, Researchers Convene to Explore Regenerative Ag appeared first on California Ag Network.

In a groundbreaking advancement in diabetes research, scientists at Sweden’s Karolinska Institutet and KTH Royal Institute of Technology have announced a refined technique for generating insulin-producing pancreatic cells from human pluripotent stem cells. This innovative procedure heralds a promising leap forward in the quest to develop effective cell replacement therapies for type 1 diabetes, a […]

Why does the same genetic mutation cause a severe brain malformation in some patients but not in others? Researchers from the MOSAIC team at the Paris Brain Institute have developed mosaic human cortical organoids carrying mutations in the DEPDC5 gene in order to model focal cortical dysplasia—a brain malformation responsible for drug-resistant epilepsy in children.

Emily Mullin / Wired: Sabi, which is developing a brain-computer interface beanie that can decode internal speech into words, emerges from stealth with backing from Khosla and others  —  California-based startup Sabi is developing a thought-to-text wearable that could usher in the cyborg future.

Olivia Poh / Bloomberg: Canva unveils Canva AI 2.0, which can generate editable layered designs from conversational prompts using Canva's foundation model built specifically for design  —  Australia's highest-profile tech unicorn is undertaking a risky transformation as it seeks to prove its relevance in the age of generative AI.

Researchers at Karolinska Institutet and KTH Royal Institute of Technology in Sweden have developed an improved method for creating insulin-producing cells from human stem cells. The results, published in Stem Cell Reports, demonstrate that these cells effectively regulate blood sugar levels in laboratory tests and can reverse diabetes in mice.

As we age, our ability to maintain healthy blood and a strong immune system gradually declines, largely because hematopoietic stem cells (HSCs), the cells responsible for producing all blood cell types, begin to lose their effectiveness. Normally, HSCs can both self-renew and generate a balanced mix of blood cells, but over time they produce fewer new cells, favor certain cells such as myeloid cells over lymphoid cells, and struggle to support a robust immune response.

Aging presents one of the most formidable challenges to human health, with the gradual decline of the hematopoietic system standing as a critical contributor to diminished immune competence and increased vulnerability to disease. Central to this decline are hematopoietic stem cells (HSCs), the foundational units responsible for lifelong blood regeneration and immune cell replenishment. Recent […]

Efforts to identify and evaluate next-generation therapeutics for pediatric brain tumors are easily stymied by the quality and availability of laboratory models for research. To address this issue, scientists at St. Jude Children's Research Hospital developed patient-derived tumor organoids and tumor organoid xenografts that accurately reflect the biologic underpinnings of embryonal brain tumors.

Turns out a lot of people would rather ask an AI about their symptoms than pay for a doctor's visit. A new survey puts some striking numbers behind that trend.

Months after a lawsuit against a San Diego health system, another class action alleges two more California health systems used Abridge’s AI tool to record patient visits without consent.

Higher torque teno virus loads during the first year after kidney transplant are associated with an increased risk for malignancies in subsequent years, a prospective study finds.

Ученые разработали новую версию CRISPR, которая способна различать ДНК опухолевых и здоровых клеток и избирательно разрезать только раковые. Работа открывает путь к более точным и безопасным методам лечения онкологических заболеваний.

The use of postmenopausal hormone replacement therapy is associated with an increased risk for glaucoma in Finnish women aged 50 years or older, according to a recent nested case-control study.

Amazon Web Services (AWS)  announced this Tuesday the launch of its AI bio tool, Amazon Bio Discovery, to accelerate the early-stage process of drug discovery in the pharmaceutical industry. The application aims to help scientists design and test novel drugs in a fast and secure way. Amazon Bio Discovery enables scientists to run complex computational […] This story continues at The Next Web

IGF-1 signaling is perhaps the most well studied mechanism of aging, with extensive work predating the modern enthusiasm for treating aging as a medical condition. Investigation of IGF-1 signaling in the context of aging was a fellow traveler to investigations of calorie restriction in the context of aging, and while these are roads that lead to a greater understanding of the evolution of aging and how pace of aging adapts to environmental circumstances, and have given rise to classes of drugs that may modestly slow aging, they are not likely to lead to any meaningful class of rejuvenation therapy. From a purely scientific point of view, the incomplete state of understanding of cellular biochemistry means that there is a lot left to learn on the […]

‘Highlander’ First Footage Unveils Henry Cavill’s ‘John Wick’-Esque Action, Rave Fight Scene and Immortal Killers  Variety'Highlander' First Look: Henry Cavill Stars in Movie Reboot  The Hollywood ReporterFirst look as Highlander reboot films at iconic Scottish castle  The Herald‘Highlander’ teaser: Henry Cavill’s ‘immortal’ warrior takes center stage; first look promises ‘epic’ journey  The Times of IndiaAmazon MGM Studios Offers First Look At Henry Cavill In New ‘Highlander’ – CinemaCon  Deadline

‘Highlander’ First Footage Unveils Henry Cavill’s ‘John Wick’-Esque Action, Rave Fight Scene and Immortal Killers  Variety'Highlander' First Look: Henry Cavill Stars in Movie Reboot  The Hollywood ReporterFirst look as Highlander reboot films at iconic Scottish castle  The Herald‘Highlander’ teaser: Henry Cavill’s ‘immortal’ warrior takes center stage; first look promises ‘epic’ journey  The Times of IndiaAmazon MGM Studios Offers First Look At Henry Cavill In New ‘Highlander’ – CinemaCon  Deadline

Feedback discovers that a conspiracy theory has formed that various celebrities have been replaced by clones, and sees just a few small problems with the idea

arXiv:2604.13795v1 Announce Type: new Abstract: Vision transformers (ViT) have been shown to allow for more flexible feature detection and can outperform convolutional neural network (CNN) when pre-trained on sufficient data. Due to their promising feature detection capabilities, we deployed ViTs for morphological classification of anaplastic large cell lymphoma (ALCL) versus classic Hodgkin lymphoma (cHL). We had previously designed a ViT model which was trained on a small dataset of 1,200 image patches in fully supervised training. That model achieved a diagnostic accuracy of 100% and an F1 score of 1.0 on the independent test set. Since fully supervised training is not a practical method due to lack of expertise resources in both the training and testing phases, we conducted a recent study on a modified approach to training data (weakly supervised training) and show that labeling training image patch automatically at the slide level of each whole-slide-image is a more practical

arXiv:2604.13783v1 Announce Type: new Abstract: Deep Learning (DL) is becoming more and more widespread in clone detection, motivated by achieving near-perfect performance for this task. In particular in case of semantic code clones, which share only limited syntax but implement the same or similar functionality, Deep Learning appears to outperform conventional tools. In this paper, we want to investigate the generalizability of DL-based clone detectors for Java. We therefore replicate and evaluate the performance of five state-of-the-art DL-based clone detectors, including Transformers like CodeBERT and single-task models like FA-AST+GMN, in a zero-shot evaluation scenario, where we train/fine-tune and evaluate on different datasets and functionalities. Our experiments demonstrate that the models' generalizability to unseen code is limited. Further analysis reveals that the conventional clone detector NiCad even outperforms the DL-based clone detectors in such a zero-shot evaluation

arXiv:2604.13427v1 Announce Type: new Abstract: Text-driven motion editing and intra-structural retargeting, where source and target share topology but may differ in bone lengths, are traditionally handled by fragmented pipelines with incompatible inputs and representations: editing relies on specialized generative steering, while retargeting is deferred to geometric post-processing. We present a unifying perspective where both tasks are cast as instances of conditional transport within a single generative framework. By leveraging recent advances in flow matching, we demonstrate that editing and retargeting are fundamentally the same generative task, distinguished only by which conditioning signal, semantic or structural, is modulated during inference. We implement this vision via a rectified-flow motion model jointly conditioned on text prompts and target skeletal structures. Our architecture extends a DiT-style transformer with per-joint tokenization and explicit joint

arXiv:2604.13381v1 Announce Type: new Abstract: Conversational generative artificial intelligence agents (or genAI chatbots) could benefit youth mental health, yet young people's perspectives remain underexplored. We examined the Mental health Intelligence Agent (Mia), a genAI chatbot originally designed for professionals in Australian youth services. Following co-design, 32 young people participated in online workshops exploring their perceptions of genAI chatbots in youth mental health and to develop recommendations for reconceptualising Mia for consumers and integrating it into services. Four themes were developed: (1) Humanising AI without dehumanising care, (2) I need to know what's under the hood, (3) Right tool, right place, right time?, and (4) Making it mine on safe ground. This study offers insights into young people's attitudes, needs, and requirements regarding genAI chatbots in youth mental health, with key implications for service integration. Additionally, by

Viral clip sparks 'Lady Gaga clone' claims online as viewers question a mysterious double seen in footage circulating across social media.

Researchers have developed a soft laser-printed scaffold made almost entirely of water that bone cells readily colonize.

In a groundbreaking advancement in regenerative medicine, a team of Japanese researchers has unveiled critical insights into the cellular architecture and stem cell dynamics within the vocal folds of the larynx, commonly known as the voice box. This organ not only orchestrates our ability to vocalize—enabling speech and song—but also manages vital physiological functions such […]

In a groundbreaking development at the intersection of cell biology and therapeutic innovation, researchers have unveiled a novel technique for targeted mitochondrial transplantation using engineered protein binders displayed on the mitochondrial surface. This cutting-edge approach enables highly selective delivery of functional mitochondria to specific cell types, overcoming longstanding challenges in mitochondrial therapy aimed at rescuing […]

'LLMs should not be trusted for patient-facing diagnostic reasoning,' boffins advise People ask AI for all kinds of advice, including the kind of questions you'd ask a physician. However, the next time you're tempted to query ChatGPT if that growth on your face is skin cancer, consider this: research shows today's leading AI models fail at early differential diagnosis in more than 8 out of 10 cases.…

Abridge is partnering with top medical journals to bring in research for doctors to refer to during visits, pitting it against medical search giant OpenEvidence as more companies push to become all-in-one platforms for doctors. ...

In a groundbreaking collaborative effort, researchers from the MRC Laboratory of Medical Sciences (LMS) and Imperial College London have unearthed a critical vulnerability within the cellular machinery of RAS-driven cancers—one that holds the promise of transforming therapeutic approaches for some of the most aggressive and treatment-resistant tumours. This discovery centers on the spliceosome, an essential […]

In the evolving landscape of boreal forest management, a recent groundbreaking study from Eastern Finland has unveiled potent strategies to harmonize commercial forestry and biodiversity conservation. Through a meticulous, landscape-scale experiment, researchers have demonstrated that the integration of prescribed burning with thoughtful tree retention fosters robust natural regeneration of pivotal boreal species, notably Scots pine […]

Cancer cells excel at evading detection, but subtle chemical differences set them apart from healthy cells. Now, a team of scientists from Wageningen University & Research and Van Andel Institute has identified a way to exploit this distinction. Using a variant of CRISPR, a modern tool for editing DNA, they distinguished tumor DNA from healthy DNA and selectively cut only the former.

Cancer cells excel at evading detection, but subtle chemical differences set them apart from healthy cells.

In a pioneering breakthrough that could revolutionize cancer therapy, researchers from Wageningen University & Research and Van Andel Institute have unveiled a novel gene-editing approach that exploits subtle chemical nuances distinguishing tumor DNA from healthy DNA. This innovative method leverages a unique variant of the CRISPR gene-editing system known as ThermoCas9 to selectively target and […]

In a groundbreaking study poised to redefine regenerative medicine, researchers have unveiled the remarkable potential of a novel gel and spray designed to accelerate tissue repair and regeneration. This cutting-edge therapeutic approach could revolutionize treatment paradigms for wounds, burns, and other tissue damage, by harnessing biologically active compounds embedded within innovative delivery systems. The team’s […]

A groundbreaking study recently published in the journal Engineering presents a pioneering approach to extending the operational lifetime of robot swarms in multi-user edge computing environments. This innovative research, conducted by experts at the Institute for Communication Systems, University of Surrey, tackles one of the most significant hurdles in the advancement of 6G-enabled swarm robotics: […]

WEDNESDAY, April 15, 2026 —For many patients with sickle cell disease, a trip to the emergency room has an unwanted side-effect: In their search for relief from agonizing pain, they are often dismissed as drug-seekers. A new study from the U...

Kickbusch’s call to reframe artificial intelligence (AI) governance through a planetary health lens is timely and compelling.1 As AI systems—particularly large language model (LLM) applications—move rapidly into use, we must ask how governance can be operationalised in healthcare practice.One of the most practical and underused levers for responsible AI is robust evaluation. In healthcare, evaluation standards have long aligned innovation with patient safety, quality, and equity. Encouragingly, this is a growing area of AI research, with emerging frameworks for assessing clinical performance,23 extensions to trial reporting such as CONSORT-AI,4 and examples of structured safety testing in real world tools.5A critical gap remains, however. Current evaluation frameworks rarely consider the computational and environmental costs of LLM powered systems, with environmental impact absent from criteria even in otherwise comprehensive approaches.6 Although recent frameworks such as FUTURE-AI7

arXiv:2604.12867v1 Announce Type: new Abstract: As agentic foundation models continue to evolve, how to further improve their performance in vertical domains has become an important challenge. To this end, building upon Tongyi DeepResearch, a powerful agentic foundation model, we focus on the Chinese medical deep search scenario and propose QuarkMedSearch, systematically exploring a full-pipeline approach spanning medical multi-hop data construction, training strategies, and evaluation benchmarks to further push and assess its performance upper bound in vertical domains. Specifically, for data synthesis, to address the scarcity of deep search training data in the medical domain, we combine a large-scale medical knowledge graph with real-time online exploration to construct long-horizon medical deep search training data; for post-training, we adopt a two-stage SFT and RL training strategy that progressively enhances the model's planning, tool invocation, and reflection capabilities

arXiv:2604.12693v1 Announce Type: new Abstract: Deep learning models often achieve expert-level accuracy in medical image classification but suffer from a critical flaw: semantic incoherence. These high-confidence mistakes that are semantically incoherent (e.g., classifying a malignant tumor as benign) fundamentally differ from acceptable errors which stem from visual ambiguity. Unlike safe, fine-grained disagreements, these fatal failures erode clinical trust. To address this, we propose Risk-Calibrated Learning, a technique that explicitly distinguishes between visual ambiguity (fine-grained errors) and catastrophic structural errors. By embedding a confusion-aware clinical severity matrix M into the optimization landscape, our method suppresses critical errors (false negatives) without requiring complex architectural changes. We validate our approach in four different imaging modalities: Brain Tumor MRI, ISIC 2018 (Dermoscopy), BreaKHis (Breast Histopathology), and SICAPv2

arXiv:2604.12108v1 Announce Type: new Abstract: Integration testing is critical for the quality and reliability of complex software systems. However, diagnosing their failures presents significant challenges due to the massive volume, unstructured nature, and heterogeneity of logs they generate. These result in a high cognitive load, low signal-to-noise ratio, and make diagnosis difficult and time-consuming. Developers complain about these difficulties consistently and report spending substantially more time diagnosing integration test failures compared to unit test failures. To address these shortcomings, we introduce Auto-Diagnose, a novel diagnosis tool that leverages LLMs to help developers efficiently determine the root cause of integration test failures. Auto-Diagnose analyzes failure logs, produces concise summaries with the most relevant log lines, and is integrated into Critique, Google's internal code review system, providing contextual and in-time assistance. Based on our

Dysfunction in the cells making up the inner lining of blood vessels, the vascular endothelium, is thought to be an important first step in the aging of the vasculature more generally, setting the stage for the development of atherosclerotic lesions, a declining capacity of smooth muscle to contract and dilate vessels in order to control blood pressure, and leakage of the blood-brain barrier, among other issues. Researchers here review the contribution of two important aspects of cellular aging to the aging of the vascular endothelium; firstly the growing number of senescent cells, and secondly the decline in mitochondrial function. These are connected, as mitochondrial dysfunction is considered to contribute to an increased pace at which cells become senescent. The vascular endothelium performs numerous regulatory functions […]

Why many Americans are turning to AI for health advice, according to recent polls  AP NewsWhere Do Americans Get Health Information, and What Do They Trust?  Pew Research CenterHave You Used A.I. Chatbots for Nutrition Advice?  The New York TimesWhy AI tests clinician trust—and how providers are responding  Modern HealthcareAmericans Turning to AI to Supplement Healthcare Visits  Gallup News