Division of Life Sciences

Ocean microbes split into distinct metabolic “niches,” each with its own carbon-processing strategy. Using global metagenomic data, researchers including Dornsife’s Naomi Levine found that these microbial differences, not just phytoplankton, shape how efficiently carbon cycles through the ocean, a major control on global climate. Most existing models only account for phytoplankton. Sharper models of ocean microbes could mean sharper climate forecasts.

Psychologist Darby Saxbe’s new book, Dad Brain, tracks first-time fathers from pregnancy through early infancy. She found measurable changes in gray matter after their child’s birth, likely reflecting the brain becoming more efficient rather than declining. The more invested a father was in caregiving, the more his brain changed, challenging the assumption that nurturing is wired only into mothers. Saxbe also found paid paternity leave lowered stress for fathers and depression in mothers, tying the science directly to family policy.

Human immune genes still carry the marks of infectious diseases our ancestors faced thousands of years ago. That evolutionary history helps explain why COVID-19 hit some people harder than others. Biologist Michael Campbell and collaborators found signs of natural selection in four immune genes across global populations, then identified rare gene variants linked to severe COVID-19. The findings connect evolutionary biology and infectious disease research in a way that could eventually point toward new strategies for preventing and treating disease.

The gut talks to the brain during meals, and that conversation shapes memory. Biologist Scott Kanoski found that nutrient-rich food triggers signals through the vagus nerve that boost a memory-forming brain chemical, likely an evolutionary tool for remembering good food sources. But chronic junk food exposure weakens that same gut-brain connection over time. Since disrupted memory signaling is an early marker of Alzheimer’s, the findings point toward vagus nerve-based therapies as a possible way to protect cognitive health.

Computational biologist Remo Rohs developed a generative AI model called Deep Predictor of Binding Specificity (DeepPBS), which is designed to predict how different proteins may bind to DNA with accuracy across different types of protein. The tool may help to reduce the time required to develop new drugs and other medical treatments.

Dean’s Professor Peter Kuhn pioneered a “liquid biopsy” that detects cancer cells through a simple blood draw instead of invasive tissue sampling. Originally built for managing late-stage disease, the test can now catch early signs of cancer, sometimes detecting the earliest interaction between a tumor and surrounding tissue. Kuhn’s team is now working to make the test accessible enough to complement mammograms and screen for cancer across the lifespan.