The Dean for Research Fund for New Ideas in the Natural Sciences supports the exploration of highly promising ideas that are at an early stage and need additional investigation prior to becoming the basis of a proposal for funding to an external agency that sponsors research. Five projects have been selected for funding this year:

Listening for the song of dark matter

Like an AM radio searching for a signal, a new tunable device could search for signatures of dark matter in the universe. Despite making up more than 80% of the universe, dark matter has never been directly detected – its presence has only been inferred through its actions on stars and gases in galaxies. Saptarshi Chaudhuri, an associate research scholar and Dicke Fellow, is designing parts for a “dark matter radio” to search for faint signals from dark-matter candidates called QCD axions. Chaudhuri and Lyman Page, the James S. McDonnell Distinguished University Professor in Physics, will build components of a high-quality tuner that can scan a range of frequencies, in a manner analogous to turning a dial on the radio, to detect these elusive particles and solve the mystery of the source of dark matter.

Aiding the search for new antibiotics

A new project will exploit the chemical warfare that microbes wage against other organisms to search for new treatments for bacterial diseases. More than three-quarters of today’s antibiotics, including penicillin, come from natural toxins that microorganisms secrete to protect themselves. The vast majority of such compounds remain unknown. Mohammad Seyedsayamdost, professor of chemistry, and his team have found evidence for the role of oxidative stress, a biochemical process, in promoting the activity of genes responsible for making these natural toxins. The team will explore how oxidative stress can aid the search for new natural toxins that could form the next generation of antibiotic drugs.

Mapping the neurons of working memory

A team will explore how neurons in the brain interact across regions to give rise to the ability to remember a phone number just long enough to punch it into a phone. Storing information in working memory is an everyday occurrence, and research indicates that working memory arises from interactions between two brain regions, one that handle sensory information (the sensory cortex) and another than handles cognitive information (the prefrontal cortex). To understand how neurons communicate across regions, Timothy Buschman, assistant professor of psychology and neuroscience, and his team will record brain activities using tiny silicon-based recording devices known as Neuropixels that capture the outputs of hundreds to thousands of neurons simultaneously. The team will use the recordings to test the hypothesis that working memory relies on the interaction of the prefrontal cortex and the sensory cortex.

Predicting Antarctic ice dynamics with deep learning

A computational approach known as deep learning could enhance the predictions of Antarctic ice dynamics associated with climate change. Sea-level rise occurs due to ice loss in regions such as Antarctica. Scientists’ understanding of how ice flows is based on decades-old laboratory studies, yet this flow law doesn’t capture processes occurring at much larger time scales such as decades and over long distances such as thousands of kilometers. Newer data from satellites could provide more accurate predictions. Ching-Yao Lai, assistant professor of geosciences and Atmospheric and Oceanic Sciences, will apply an approach called physics-informed deep learning to find the fundamental flow law of ice from satellite data and build better predictions of ice dynamics in a changing climate.

Tracking RNA in living cells

Researchers will develop new, versatile tools to track the movement of RNA, the cell’s messenger delivery system, and to remove selected RNAs from cells. A chemical cousin of DNA, RNA moves about the cell carrying instructions to the precise locations where proteins need to be made. A team led by Cameron Myhrvold, assistant professor of molecular biology and a Class of 2011 alumnus, and Elizabeth Gavis, the Damon B. Pfeiffer Professor in the Life Sciences and professor of molecular biology, will explore where and why RNAs travel. They’ll develop tools based on the CRISPR-Cas13 system, a cousin of the CRISPR DNA-editing technique, that targets RNA to illuminate the locations of many different RNAs in the cell and probe their functions at these locations. The team hopes to uncover roles for transported RNAs in the behavior of cells in a developing embryo.