Toggle light / dark theme

Get the latest international news and world events from around the world.

Log in for authorized contributors

Beyond Transformers: Why Graph Neural Networks Are the Next Frontier in AI

In contemporary artificial intelligence, transformers are everywhere, changing the way we do everything from natural language processing to computer vision. People have rushed to play with GPT-4 and other AI text models built on top of Transformer architectures because the machines are capable of solving problems they previously couldn’t, riffing stories, code or poetry, creating images from sentences, even speaking like Turing test-worthy humans. But as artificial intelligence improves, researchers have found that a grid-based or sequential approach to data has increasingly stringent constraints. And there’s a new AI technology that holds the potential of unraveling the mysteries that our complex and interconnected world holds around us: Graph Neural Networks (GNNs).

Gene signature an early warning system for aggressive pancreatic cancer

Precancerous cells must adapt to and overcome cellular stress and inflammation in order to progress and form malignant tumors. Now, researchers have identified a link between stress and inflammation and pancreatic ductal adenocarcinoma (PDAC), one of the most aggressive and lethal types of cancer. The findings could serve as an early warning system for the disease, leading to the detection of PDAC before it becomes life-threatening.

Previous studies have shown that inflammation and cellular stress activate a protein called STAT3 — short for signal transducer and activator of transcription 3 — in pancreas cells, promoting tumor initiation, adaptation to stress and resistance to treatment. How STAT3 accomplishes this has not been understood until now.

In the current study, the researchers discovered that in some cancer cells, STAT3 is able to activate specific genes critical for adaptation to stress and inflammation. They found:

Scientists Find 2 Existing Drugs Can Reverse Alzheimer’s Brain Damage in Mice

In efforts to beat Alzheimer’s disease, researchers are looking at existing drugs that could tackle the condition, and a new study identifies two promising candidates that are currently used to treat cancer.

Already approved by regulators in the US – meaning potential clinical trials for Alzheimer’s could start sooner – the drugs are letrozole (usually used to treat breast cancer) and irinotecan (usually used to treat colon and lung cancer).

Researchers from the University of California, San Francisco (UCSF) and Gladstone Institutes started by looking at how Alzheimer’s altered gene expression in the brain.

AI turns immune cells into precision cancer killers—in just weeks

A breakthrough AI system is revolutionizing cancer immunotherapy by enabling scientists to design protein-based keys that train a patient s immune cells to attack cancer with extreme precision. This method, capable of reducing development time from years to weeks, was successfully tested on known and patient-specific tumor targets. Using virtual safety screenings to avoid harmful side effects, the platform represents a leap forward in personalized medicine.

Brain peptide ODN reduces hunger and boosts glucose regulation in rat study

University of Pennsylvania and Syracuse University scientists have discovered that a hindbrain-derived peptide, octadecaneuropeptide (ODN), can suppress appetite and improve glucose regulation without causing nausea or vomiting. Results suggest a glia-to-neuron signaling axis in the dorsal vagal complex that may be harnessed for treating obesity and type 2 diabetes.

Glial cells in the brainstem produce ODN, a signaling peptide whose physiological role in energy homeostasis has remained obscure. Researchers now find that directly activating this peptide system in the hindbrain induces weight loss, enhances glucose disposal, and lowers in obese animals.

Unlike existing therapies targeting GLP-1 receptors, ODN achieves these effects without triggering nausea-related behaviors or emesis in vomiting-competent models.