(Cell Reports 45, 116734; January 27, 2026)
When this paper was published on December 18, 2025, Figure S2 inadvertently omitted panel C. The article has since been updated online with the correct, complete figure. For readers’ convenience, both the original and corrected figures can be seen below. The authors apologize for the error.
Clinically available KRAS inhibitors mainly target G12C, which is rare in PDAC and often acquires resistance. Oncogenic KRAS inactivates RB1 via CDK4/6, while RB1 mutation is rare. Thus, CDK4/6 inhibition offers an indirect strategy to counter KRAS-driven malignancy without direct KRAS targeting.
Virtually all pancreatic ductal adenocarcinomas (PDACs) are initiated by activating mutations in the oncogene KRAS, which occur in multiple distinct allelic forms. Although considerable efforts have led to the development of inhibitors targeting specific mutant KRAS proteins, the only agents currently approved for clinical use selectively target the KRASG12C variant. However, KRASG12C mutations are exceedingly rare in pancreatic cancer.
Furthermore, in patients with KRASG12C-mutant pancreatic cancer, treatment with KRASG12C inhibitors has shown only modest clinical benefit, comparable to that of conventional chemotherapeutic regimens, and even in cases with an initial objective response, acquired resistance almost invariably emerges within a limited time frame.
People with autism may be up to six times more likely to develop Parkinson’s disease in later life. New research offers a potential explanation based on the role of transporter molecules that recycle unused dopamine in the brain.
Dopamine is a neurotransmitter crucial for managing movement and executive functions, and for reinforcing behavior. It’s well known that Parkinson’s is characterized by a drop in dopamine levels, while disruptions in the transport of the chemical have also been linked to autism.
With that context, researchers led by a team from the University of Missouri in the US took a novel approach using a technology known as a DaT SPECT scan, which is typically used to diagnose Parkinson’s in much older people.
UTICalc demonstrated strong diagnostic performance for UTI in febrile children aged 2 to 24 months, supporting its use as an evidence-based adjunct in emergency department patient assessment.
This multicenter diagnostic study prospectively provides the final stage of external validation for UTICalc in young febrile children. We found strong discrimination for the model, which was improved when incorporating dipstick data. This study supports integration into clinical care as a tool for diagnostic stewardship in pediatric care.
Decision curve analysis supported the utility of both UTICalc models across relevant thresholds. However, clinician judgment demonstrated higher sensitivity in the high-volume tertiary centers where this study was conducted, reflecting pediatric emergency medicine practitioner expertise. Because most children presenting for acute care are evaluated outside tertiary centers,27 possibly with limited pediatric expertise and uncertain follow-up, using a 5% risk threshold—despite reducing testing—may not be ideal due to reduced sensitivity. UTICalc may serve as a useful adjunct for clinicians with less pediatric experience or in cases of diagnostic uncertainty, with lower risk thresholds (eg, 2%) potentially more appropriate for patients with persistent symptoms or anticipated barriers to follow-up.
The observed UTI prevalence of 4% in our cohort is consistent with previously reported risk estimates in this population, which ranges from 3% to 11%.2,28,29 Model performance in our study was comparable with the original derivation and validation study published in 2018, which reported an AUROC of 0.80 in the clinical model and 0.97 in the clinical and dipstick model. Low PPV in our sample is reflective of low disease prevalence.
Could future gasoline come from thin air and sunlight instead of oil wells? A team of Chinese scientists has unveiled a lab system that imitates plant photosynthesis to turn carbon dioxide and water into gasoline building blocks using only sunlight. Their work hints at a way to recycle a major greenhouse gas while still using existing engines and fuel infrastructure.
In an artificial photosynthesis study, the researchers report a “charge reservoir” material that stores solar energy as electrical charge, then delivers it on demand to drive reactions. The system converts carbon dioxide into carbon monoxide, a key building block for synthetic fuels, and uses water as its only electron source instead of extra helper chemicals.
Although still a lab device, the setup works under natural sunlight and is meant to connect renewable energy to industry and transport.
Please see my latest Forbes article: The Rapid Trajectory of Artificial Intelligence: From Machine Learning Foundations to Generative Creativity, Agentic Autonomy, Human Augmentation, Neuromorphic Intelligence, and the Cyborg Horizon.
Thanks and have a great weekend!
#artificialintelligence #tech #ai #future @forbes
Artificial intelligence continues to evolve at an accelerating pace, transitioning from narrow, data-driven tools to systems capable of reasoning, and autonomous action.
Come with me if you want to live. – The Terminator.
‘Close enough’ only counts in horseshoes and hand grenades. – Traditional.
Cobalt and nickel are vital components for batteries, superalloys and catalysts, used in technologies ranging from smartphones to jet engines. But when it comes to recycling, they are notoriously difficult to separate because they are chemically nearly identical. To solve this, a team led by scientists at Johns Hopkins University in the United States has developed a cleaner and cheaper way to extract these elements. And it is thanks in part to grapes.
