For decades, scientists have been baffled by two enormous, enigmatic structures buried deep inside Earth with features so vast and unusual that they defy conventional models of planetary evolution.

Now, a study published in Nature Geoscience by Rutgers geodynamicist Yoshinori Miyazaki in combination with collaborators offers a striking new explanation for these anomalies and their role in shaping Earth’s ability to support life.

Nematologists have long wondered how the the Northern root-knot nematode (Meloidogyne hapla) is able to infect such a wide range of plants, from carrots to trees. Now an international team has gained insight into how the DNA of this parasitic worm facilitates its success. 

Research led by Megan Matthews has solved a 70-year old medical mystery: how one of the oldest FDA-approved drugs for hypertension and preeclampsia, hydralazine, actually works. In the process, they revealed that it can also disarm one of the deadliest brain cancers. “Once prescribed as one of the first vasodilators for hypertension, hydralazine remains a first-line treatment for preeclampsia, a pregnancy-related hypertensive disorder that accounts for 5 to 15 percent of maternal deaths worldwide,” says Kyosuke Shishikura. “Yet for decades, no one knew exactly how the drug worked. Now that we understand its mechanism, we can start designing safer, more selective versions to target both vascular and cancer diseases.”

Earthquake faults deep in the Earth can glue themselves back together following a seismic event, according to a new study led by researchers at UC Davis. The work adds a new factor to our understanding of the behavior of faults that can give rise to major earthquakes. 

Digital touch has lagged behind progress in audio and video. New device, called VoxeLite, is first to accurately match the complex, detailed sensing ability of a human fingertip. Weighing less than a gram, VoxeLite enables wearers to feel virtual textures and patterns on smooth surfaces. New tech could give tactile feedback to users with vision impairments or enhance virtual reality systems.

The Context-Guided Segmentation Network (CGS-Net) developed by University of Maine researchers introduces a deep learning architecture designed to interpret microscopic images of tissue with greater precision than conventional AI models. Powered by a dual-encoder model that mirrors the workflow of a pathologist examining a slide, one branch of the network processes a high-resolution image patch to capture cell-level details, while the other examines a lower-resolution patch encompassing the surrounding tissue. A system of interconnected encoders and decoders uses data from both the high and low resolution images for a complete analysis. 

A team of Iowa State researchers is establishing a year-long program to recruit and train students for the semiconductor industry. It's all about giving students a head start on a semiconductor career and building a U.S. chip workforce.