City of Hope and UC Berkeley Researchers Teach AI to Spot Cancer Risk by Squeezing Individual Breast Cells Scientists have developed a groundbreaking method to assess breast cancer risk by analyzing the mechanical properties of individual breast cells. Researchers at City of Hope, a leading cancer research and treatment organization, and the University of California, Berkeley, have created a microfluidic platform called MechanoAge, which uses machine learning to identify individuals at higher risk for breast cancer based on how their cells respond to stress. The study, published in eBioMedicine, marks a significant advancement in personalized cancer risk assessment. The technique, known as mechano-node pore sensing (Mechano-NPS), involves squeezing individual breast epithelial cells through a narrow channel to measure their deformation, recovery, and behavior under pressure. By analyzing these physical responses, the platform generates a quantifiable "mechanical age" for cells, which correlates with cancer risk. This approach addresses a critical gap in risk assessment, as over 90% of women lack known genetic predispositions or family histories of breast cancer, making traditional methods less effective. Dr. Mark LaBarge, a professor at City of Hope, emphasized the importance of this tool for women without genetic risk factors. "For everyone else, you’re left wondering, ‘Am I at high risk?’" he said. "By translating physical changes in cells into quantifiable data, this tool gives women something tangible to discuss with their doctors—not just risk estimates, but evidence drawn directly from their own cells." The platform’s machine learning algorithm identifies cells exhibiting signs of accelerated aging, assigning a risk score based on mechanical and physical properties.#university_of_california_berkeley #mechanoage #city_of_hope #dr_mark_labarge #dr_lydia_sohn
City of Hope and UC Berkeley Scientists Develop AI-Driven Platform to Assess Breast Cancer Risk via Cellular Mechanics Scientists at City of Hope and the University of California, Berkeley, have developed a groundbreaking microfluidic platform that uses artificial intelligence to assess breast cancer risk by analyzing the mechanical properties of single breast epithelial cells. The innovation, published in The Lancet’s eBioMedicine, introduces a novel method to evaluate cellular aging and stress resilience, offering a direct biophysical measure of cancer susceptibility. This technology marks a significant departure from traditional risk assessment tools, which have historically relied on genetic factors and indirect methods like mammographic breast density. The platform applies mechanical stress to individual cells by squeezing them through narrow microfluidic channels, mimicking biomechanical stressors. By measuring how quickly cells deform and recover their shape, researchers can quantify their "mechanical age"—a concept borrowed from material engineering that assesses wear and fatigue in metals and polymers. This approach reveals subtle differences in cellular behavior that correlate with heightened cancer risk, even in individuals without known genetic predispositions. For example, some younger women’s cells exhibited stiffness and prolonged recovery times, indicating advanced mechanical aging despite their chronological age. Traditional methods, such as genetic testing for mutations like BRCA1/BRCA2, account for only about 6% of breast cancer cases. For the remaining 94%, risk stratification has been imprecise, often leading to over-diagnosis or missed early warnings. The MechanoAge platform addresses this gap by providing a direct, cell-level assessment.#national_institutes_of_health #university_of_california_berkeley #city_of_hope #the_lancet_ebmedicine #mechanoage

Stanford scientists discover "natural Ozempic" without side effects Scientists at Stanford Medicine have identified a naturally occurring molecule that mimics the weight loss effects of semaglutide, the drug known as Ozempic, but with fewer side effects. The molecule, named BRP, was discovered using artificial intelligence and appears to act directly on the brain’s appetite-control center, reducing food intake and promoting fat loss without causing nausea, constipation, or muscle loss. The findings, published in Nature, could pave the way for more targeted and safer weight loss treatments. The discovery relied on a computational tool called Peptide Predictor, which analyzed human protein-coding genes to identify potential peptides derived from prohormones. These molecules, initially inactive, can be split into smaller fragments that function as hormones. The team focused on an enzyme, prohormone convertase 1/3, linked to obesity, and narrowed their search to proteins secreted outside cells, which are key features of hormones. This process identified 373 prohormones, from which the algorithm predicted 2,683 possible peptides. Researchers tested 100 of these, including GLP-1, a hormone known to regulate appetite and blood sugar. Among the tested peptides, a 12-amino-acid fragment, named BRP (short for BRINP2-related-peptide), showed the most potent effects. In lab-grown brain cells, BRP significantly increased neuronal activity, tenfold higher than GLP-1. Animal studies further demonstrated its efficacy: in lean mice and minipigs, a single injection reduced food intake by up to 50% within an hour. Over 14 days, obese mice treated with BRP lost an average of 3 grams of body weight, primarily fat, while untreated animals gained similar amounts.#stanford_medicine #brp #peptide_predictor #merrifield_therapeutics #university_of_california_berkeley
