Archived Intelligence
A new machine learning framework accurately simulates complex chemical behaviours by integrating long-range physical interactions. Trained on elements across the periodic table, this foundation model balances computational speed with the precision needed for advanced materials science.
A systematic review of TA-65 confirms it successfully elongates telomeres in older adults yet fails to improve physical strength or inflammation levels. The study highlights a disconnect between cellular biomarkers and actual functional health, alongside evidence of inflated results in industry-funded trials.
Antimicrobial resistance is a growing global threat requiring advanced surveillance strategies. A new review highlights how artificial intelligence, ranging from deep learning to large language models, is revolutionising the detection of resistance proteins and the design of novel inhibitors to block them.
Researchers have identified a specific genetic mutation in the MIDEAS gene that causes a multisystem disorder characterised by developmental delays and physical anomalies. This mutation disables a molecular safety mechanism, causing the MiDAC protein complex to become dangerously hyperactive.
Researchers have identified that SREBF2, a key controller of cholesterol metabolism, acts as a significant driver of disease progression in acute myeloid leukaemia. High levels of this protein correlate with poor patient survival, marking it as a vital new tool for prognosis.
New research links the strength of transnational climate policies directly to the ecological stability of moon jellyfish in the Baltic Sea. By analysing two decades of data, scientists demonstrated that robust international regulations help mitigate carbon emissions and nutrient pollution, fostering a more resilient marine environment.
Early detection of oral cancer is often hampered by the limited accuracy of current screening methods, leading to diagnoses at dangerous, advanced stages. By utilising a specifically created dataset of tongue portraits, researchers demonstrated that the VGG16 deep learning model can identify pre-cancerous lesions with a validation accuracy of over 89 per cent.
Cyproterone acetate is a common drug for treating hirsutism, a condition causing excessive hair growth in women. A major review finds that while effective, it generally performs similarly to alternative medications like spironolactone, though weak evidence from small studies limits definitive conclusions on safety and superiority.
Researchers utilised a high-throughput platform to analyse the genetic 'fingerprint' of a rare form of autism called 19q12 ASD. By screening existing drugs, they discovered that entrectinib could reverse the disease's genetic signature, leading to successful off-label treatment in a patient.
Standard CRISPR/Cas9 editing is limited by its preference for specific DNA markers, leaving parts of the yeast genome inaccessible. A new study introduces an optimised CRISPR/Cas12a system that targets thymine-rich areas, allowing scientists to edit previously unreachable genetic sequences with high precision.
Researchers have successfully used CRISPR base editing to correct a specific mutation in the BAP1 gene found in kidney cancer cells. By repairing this genetic error, they restored the cells' sensitivity to anoikis—a natural self-destruct mechanism triggered when cells detach from their surroundings.
A systematic review of 34 randomised clinical trials suggests that probiotic supplements may improve cognitive function, particularly after 12 weeks of use. However, while test scores for memory and fluency showed promise, the overall certainty of the evidence remains low.
New research reveals that social isolation during adolescence can physically remodel the brain circuits responsible for processing touch in mice. This structural change, driven by stress hormones, weakens individual neural connections and impairs the ability to distinguish fine textures.
Following the severe 2024 mass bleaching event, coral reefs in southwest Madagascar's Mahafaly Seascape demonstrated unexpectedly high recovery rates. The study reveals that despite significant thermal stress, mortality remained low, suggesting these understudied ecosystems may serve as resilient sanctuaries against climate change.
A new study along the Tuticorin coast reveals that vital mangrove ecosystems are being suffocated by plastic waste, with trapped litter reaching alarming densities. Researchers found that over half of the debris is plastic, originating nearly equally from land and sea sources, posing significant ecological risks.
Scientists have engineered a robust catalyst that converts carbon dioxide into valuable chemicals at ambient pressure. By trapping palladium nanoparticles within a nitrogen-rich scaffold, the system prevents degradation and achieves near-perfect selectivity.
Researchers have developed ResTab Net, a new AI system that combines tissue images and protein data to diagnose breast cancer with 98.56% accuracy. Unlike traditional 'black box' models, this system uses explainable AI techniques to show clinicians exactly how it reaches its diagnostic conclusions.
Researchers have developed NeuroAgeFusionNet, a hybrid deep learning framework that combines three distinct AI architectures to estimate biological brain age from MRI scans. By accurately capturing complex structural connectivity patterns, this tool achieves state-of-the-art performance for early neurodegenerative disease detection.
Scientists have upgraded lithography-based 3D printing by replacing smelly, unstable thiols with readily available alcohols. Using light-activated catalysts, this 'oxa-Michael' reaction creates tough, uniform polymer networks ideal for high-temperature manufacturing.
Researchers have achieved ultra-high quality factors in nanophotonic resonators by significantly reducing the speed of light within the device. By utilising spectral hole burning in erbium-doped lithium niobate, they enhanced performance by nearly 1,000 times, paving the way for advanced quantum communication and sensing.
New research demonstrates that hearing aids equipped with deep learning algorithms significantly reduce the mental exertion required to listen in noise. By measuring brain oxygen levels, scientists found that these devices allow the prefrontal cortex to work less intensely while improving speech understanding.
Scientists have engineered a highly durable copper-nickel catalyst that resists degradation during the conversion of carbon dioxide into carbon monoxide. By anchoring nanoparticles with hydroxyl groups and harnessing light energy, the material maintains over 99 per cent of its efficiency for nearly a month.
Scientists are unlocking the secrets of 'inherently chiral macrocycles', unique ring-shaped molecules that lose their handedness if broken. Previously too difficult to synthesise, new methods are finally allowing researchers to exploit their potential for advanced materials and catalysis.
Researchers have analysed the protein cargo of tiny particles called extracellular vesicles released by neurons during cortical spreading depolarization. The findings reveal that neurons undergo complex structural and metabolic adaptations during migraine events, potentially offering a new way to track the condition via biomarkers.
Scientists have developed a new technique to characterise the proteins within specific, microscopic brain regions in the mouse hippocampus. This study reveals unexpected diversity in synaptic proteins and highlights how individual neurons regulate glutamate receptors to fine-tune communication.
Researchers have developed a sophisticated modelling framework to simulate how repetitive magnetic stimulation physically alters brain connections. By replicating real-world experiments, this tool predicts how specific frequencies and locations strengthen synapses, aiding the design of better therapeutic protocols.