Research using machine learning to solve scientific problems. Covers AI quality control, sea-surface temperature estimation, label-free cell analysis, and cyber defense.
Researchers at the Netherlands Cancer Institute developed a deep learning model PARM that predicts gene regulation with unprecedented accuracy. The model enables the prediction of functional impact of regulatory mutations in specific cell types, opening new paths for cancer diagnostics and patient stratification.
Researchers at Queen Mary University of London have developed an AI system called Auto MV that can generate complete music videos directly from full-length songs. The system uses a multi-agent approach to analyze the musical structure and lyrics, plan scenes, and generate images and video clips.
Researchers at MIT developed a generative AI model called DiffSyn that suggests promising synthesis routes for complex materials like zeolites. By using this model, scientists can test millions of theoretical materials in under a minute, accelerating the materials discovery process.
The University of Ottawa has launched the Ottawa Medical Artificial Intelligence Research Institute (OMARI), a center for research, education, and innovation in medical Artificial Intelligence. Led by Dr. Khaled El Emam, OMARI aims to facilitate cross-cutting collaborations and sharpen the university's competitive edge in AI-driven hea...
Researchers at Boston Children's Hospital have developed a novel mass spectrometry tool called FLEXITau to analyze brain tissue from 203 patients with various tauopathies. The study identified 145 post-translational modifications and 195 cleavage sites across tau, providing a precise molecular roadmap for diagnostics and drug development.
The portable optical sensor uses machine learning to analyze spectral patterns and estimate grape ripeness directly on the vine. This technology promises significant business benefits for winemakers by providing non-destructive, real-time insight into grape ripeness, reducing labor and time required for sampling and analysis.
The FIU-Florida Heart Research Foundation Center for Innovation in Cardiovascular Health brings together experts in various fields to uncover the fundamental drivers of cardiovascular disease. The center will focus on AI-powered diagnostics, regenerative tissue technologies, and workforce development.
Researchers developed a machine learning framework that accurately predicts and optimizes biochar production from algae, identifying temperature as the dominant control on biochar yield. The model achieved strong agreement with experimental results and was able to pinpoint key factors influencing biochar production.
A new study using AI-powered BioBERT model accurately identifies fungal lifestyles, switching between helpful partner for plants to aggressive decomposers. The tool has nearly 90% accuracy and can scan thousands of papers in minutes, flagging species that may switch roles.
A new AI-driven model predicts patient risk of suffering a major adverse cardiac event (MACE) with higher accuracy than traditional predictive models. This improvement in prognostic risk assessment could help personalize individual care plans and maintain quality of life for patients.
Researchers have developed a new AI tool called VASCilia that provides unprecedented 3D views of cochlear hair cells. The tool accelerates the imaging process by 50-fold, allowing scientists to analyze cells with greater precision and accuracy. This advancement offers new insights into hearing loss caused by damaged inner ear hair cells.
Researchers have developed a passive, solar-powered orbital data center that can scale AI computing and reduce environmental impact. The system leverages decades of research on 'tethers' and could host thousands of computing nodes to replicate terrestrial data centers.
A recent study from Binghamton University School of Management reveals that focusing on human-robot collaboration can generate additional economic value and improve a company's ability to capture a greater share of the competitive market. By leveraging robots in collaborative settings, organizations can foster a positive sense of commi...
Researchers at Chungnam National University have developed an AI model that uses deep learning to predict stable defect configurations in materials. The model, trained on data generated by conventional simulations, can generate results in milliseconds rather than hours, accelerating the material design process.
The Institute of Science and Technology Austria (ISTA) has received a significant donation to advance trustworthy AI technology. The €5 million gift from Garrett Camp will support fundamental research in artificial intelligence, focusing on interdisciplinary collaboration and long-term impact.
The Center for AI Responsibility and Research will build upon Binghamton research advancing AI for the public good. The new center is supported by a $30 million commitment from Tom Secunda and $25 million from Gov. Hochul and the NY State Legislature.
Researchers found that complexity, including ground truth, real-world complexity, and stakeholder involvement, are key factors in reducing AI biases. By accounting for these complexities, developers can create more fair AI models.
A large-scale study reveals that generative AI models have reached the threshold of average human creativity, but the most creative individuals still outperform even the best AI systems. The study also highlights the importance of human guidance and parameterization in modulating AI creativity.
Researchers created a computational model that combines physiological signals, sensory input, and word information to construct human emotions. The model achieved an agreement rate of about 75% when compared to participants' self-reported emotional evaluations.
Researchers developed an autonomous AI network technology that enables multiple AI systems to spontaneously collaborate and share knowledge to discover new materials. This collaboration improves exploration efficiency and paves the way for the creation of numerous new materials.
Researchers developed an AI-based system that accurately detects whip sounds in horse racing, achieving detection rates of up to 70% in audio data. The system's ability to process audio in real-time and its reliance on high-frequency components make it a promising tool for improving animal welfare and fair competition.
The review synthesizes over a decade of research on ammonia inhibition in anaerobic digestion, highlighting emerging tools to stabilize biogas production from nitrogen-rich wastes. Biological and material-based solutions, digital technologies, and synthetic biology are discussed as potential mitigations.
A chemist proposes a framework for shared model proteins to improve reproducibility and coordination in protein science. The proposal includes five widely used proteins and aims to establish minimal reporting requirements and curated reference datasets.
LMU researchers created a tool that combines automated chemical synthesis, high-throughput characterization, and data-driven modeling to control nanocrystal growth. The Synthesizer platform enables precise predictions of material properties, such as color, brightness, or stability, for applications like LEDs, solar cells, and sensors.
Researchers from Korea University report a breakthrough in reviving an abandoned depression drug target by redesigning the molecular structure of neurokinin-1 receptor antagonists. New compounds exhibiting antidepressant-like effects have been identified, reducing depressive-like behavior and brain inflammation in mice.
Engineers at the University of Pennsylvania have discovered that foams exhibit internal motion resembling deep learning in AI systems. The study suggests a common mathematical principle underlying both foams and AI training, with implications for designing adaptive materials and understanding biological structures.
Researchers have decoded the logic of microRNA strand selection using AI, revealing a conserved and programmable mechanism governing gene regulation. The study found that this decision follows conserved rules rather than chance, with mammalian microRNAs showing a strong bias towards a single dominant strand.
Researchers developed a new technique called CLASSIC that enables large-scale testing of complex DNA circuits in human cells. The approach uses artificial intelligence and machine learning to analyze vast numbers of complete circuits at once, providing scientists with a clearer picture of the rules governing genetic part behavior.
Researchers at the University of Groningen developed an atomistic model that predicts the driving force for microstructural twinning in shape memory alloys. This discovery can lead to the creation of new crystalline materials with improved reversible deformations, vibration damping, and impact absorption.
A deep learning model trained on stage II colorectal cancer whole slide images accurately identified features linked to recurrence risk. The study found the model surpassed clinical prognostic parameters in predicting patient outcomes.
A team of researchers developed an AI algorithm capable of classifying complex topological phases of matter without relying on traditional mathematical tools. The breakthrough tackles the notoriously difficult realm of non-Hermitian systems and suggests that AI can surpass human capabilities in certain domains of abstract reasoning.
Researchers at Colorado State University used AI to modify antibodies into stable intrabodies that can visualize histone modifications in real-time. This allows for better understanding of gene expression and its relationship with cancer and other disorders. The team created 19 new antibody-based probes with a 70% success rate, signifi...
A machine-learning-based algorithm developed by Tokyo Metropolitan University researchers can accurately count sister chromatid exchanges (SCEs) in chromosomes, giving a more objective measurement. The accuracy rate is 84%, which could help diagnose disorders like Bloom syndrome with greater consistency.
Artificial intelligence is transforming water quality management by detecting subtle biological changes and triggering timely responses. AI-based sensors can analyze signals directly in the field for near real-time water quality assessment, predicting harmful algal blooms and pathogen outbreaks.
The Mount Sinai Health System has launched an AI-powered clinical trial-matching platform, PRISM, to help connect patients with life-saving trials. The platform uses machine learning to match patients with trials based on their electronic health records and diagnosis.
A modular machine learning framework using LoRA fine-tuning can accurately predict single organic reactions and handle multiple reaction types. The model demonstrates high prediction accuracy comparable to expert experience, achieving results on classic reaction datasets and showing strong adaptability.
A new AI tool, CattleFever, uses artificial intelligence and thermal cameras to estimate cattle body temperature from a photo. The system can automatically determine an animal's body temperature within 1 degree of the reading from a thermometer.
Researchers at Princeton University developed a machine learning tool to predict the stability of MOF structures, allowing for faster discovery of advanced materials. The tool achieved accurate predictions 97% of the time and could lead to breakthroughs in battery chemistry, carbon capture, and clean water access.
Researchers at Mount Sinai have developed an AI-powered ECG analysis tool that shows promise in detecting Chronic Obstructive Pulmonary Disease (COPD) early. The model achieved high accuracy rates across diverse populations, including a subgroup with irregular heartbeat and smoking exposure.
Researchers combined Dynamic Optical Contrast Imaging (DOCI) with machine learning to identify thyroid cancer during surgery. The AI analysis framework accurately classified samples across three categories and generated tumor probability maps for precise location identification.
Jundong Li, an associate professor at the University of Virginia, has received the 2025 Tao Li Award for his significant contributions to data mining and machine learning. His research focuses on developing models that can extract actionable insights from structured data, particularly graphs.
A study demonstrates how optimized genomic prediction models can reliably forecast complex agronomic traits in rapeseed. The research achieves high prediction accuracy for multiple economically important traits, particularly flowering time and yield components.
A comprehensive review outlines how ammonia disrupts biogas production and identifies practical strategies to prevent system failure. Emerging solutions include the integration of artificial intelligence and machine learning into digester monitoring.
A study published in EMBO Molecular Medicine has identified a combination of statins and phenothiazines that shows promise in treating aggressive neuroblastoma. The drug combination was found to impede tumour growth and improve survival rates in laboratory trials with mice.
The new framework reduces processing time by an average of 87.7% compared to the state-of-the-art approach, achieving cache hit rates between 70% and 90%. It demonstrates linear scalability, handling high query loads efficiently.
HSE economists found that AI models like ChatGPT and Claude tend to play 'too smart' and lose in strategic thinking games by assuming a higher level of logic in people than is actually present. The study replicated results from previous human participant experiments, showing LLMs adapt to opponents with varying levels of sophistication.
A newly developed AI model accurately predicts blood loss in patients undergoing high-volume liposuction, showing excellent agreement with estimated volumes. The model's 94% accuracy may help make surgery 'smarter and safer' by informing perioperative management and reducing the risk of complications.
A new study analyzed over 460,000 scientific abstracts to identify key themes, trends, and research gaps in aging research. The study found a growing separation between basic biological studies and clinical research, highlighting the need for integration to translate laboratory discoveries into medical applications.
The new program focuses on advancing foundational research in AI, including innovation in language models and algorithmic efficiency. Google will support research grants, scholarships for students, and educational initiatives at the TAU Center.
Researchers developed RNACOREX, a new open-source software tool that identifies gene regulation networks in cancer. The tool analyzes thousands of molecules simultaneously to detect key interactions, providing an interpretable molecular map that improves understanding of tumors.
Researchers developed a tool to quickly identify resistant strains of S. aureus using genomic profiles and machine-learning models. The approach is based on gene-content information rather than highly detailed genomic profiles, making it more practical for real-life clinical contexts.
Researchers developed a machine learning framework, DeePKS, to improve density functional theory calculations for CO adsorption on metal surfaces. The framework achieves near-hybrid functional accuracy with high efficiency, enabling the exploration of complex catalytic systems.
Researchers from HIRI analyzed over 24,000 bacterial genomes to demonstrate that common machine learning methods can lead to biased results. They found that conventional approaches may fail to capture true resistance signals and limit accuracy in new strains.
A new AI framework uncovers simple, understandable rules governing complex dynamics in nature and technology. The AI generates equations that accurately describe complex systems, revealing hidden variables that govern their behavior. This approach offers scientists a new way to leverage AI for understanding complex systems.
A new study from the University of East Anglia finds that AI tools excel at speed and clarity, but human translators remain crucial for tone, cultural nuance, and natural sound. Generative AI can handle parts of video translation, but consumers still notice when something feels off.
A novel machine learning framework predicts protein binding on gold nanoclusters, revealing chemical principles governing biomolecule–gold interactions. The model enables scalable design of effective nanomaterials for biomedical applications.
Researchers developed an electronic nose that can detect and identify two common indoor mold species using nanowires. The e-nose measures changes in electrical resistance to gas molecules interacting with a sensing material, proving its potential for fast and objective monitoring of indoor air quality.
Researchers used machine learning to combine clinical data with biological markers, achieving 89.58% accuracy in predicting patient survival. The study identified key markers such as E2F8, WDR77, and hsa-miR-495-3p, which were associated with tumor growth and cancer development.
Researchers developed a transfer learning-enabled framework to predict mechanical properties of particle-reinforced aluminum matrix composites. The model achieves high predictive accuracy, addressing the challenges of limited data availability in traditional machine learning algorithms.
Researchers have launched a new multimodal medical dataset, Observer, capturing anonymized, real-time interactions between patients and clinicians. The dataset links video, audio, transcripts, and electronic health records to study subtleties like body language and environmental factors affecting care.