Many Rice students discover their path through a memorable professor or class. For senior Sophia Lyu, that turning point came midway through college. She arrived at Rice excited to explore different majors, initially selecting statistics, but exploring options for a double major. Her curiosity was ignited during a sophomore-year organic chemistry class taught by Kasey Leigh Yearty.
“I took Organic Chemistry in the spring semester of my sophomore year and really enjoyed that class,” Sophia recalled. “I remember thinking, ‘Oh, maybe I should take more chemistry classes.’”
That decision led her to the group of theoretical chemist Anatoly Kolomeisky, a pivotal experience. “He is so amazing. He made this major fantastic,” she said. Under his guidance, she found a way to combine her background in statistics with biophysical chemistry.
“Dr. Kolomeisky’s group was interesting to me because he’s doing biophysical chemistry with a lot of mathematical derivations and computer simulations, which is very statistical stuff. It is an interesting topic to me, and I can also address these problems,” she said. Sophia published two papers during her junior year, each addressing a critical biological transition by using mathematics to capture the inherent randomness of biological processes.
Her first project examined the timing of cell lysis in bacteria. Bacterial viruses infect cells and trigger the production of a type of protein called holins. When enough holin proteins accumulate in the bacterial membrane, the membrane becomes permeable and the cell is destroyed.
Kolomeisky’s group had an existing model explaining how lysis depends on the balance between the rate at which holins insert into the membrane and the rate at which they leave due to rupture. Sophia expanded this framework by adding additional layers of biological complexity, including the random production of mRNA and the transcriptional bursts that naturally occur in cells. “I included more factors into this model to make it closer to reality,” she explained.
Her results showed that although these random processes do not affect the threshold-like behavior of lysis, they did shift the precise values of the threshold and broaden the distribution of outcomes. This suggests that even with this added biological noise, cell lysis remains a robust phenomenon. This insight has important implications for bacteriophage therapy as a potential alternative to antibiotics.
Her second project focused on a very different biological question: why menopause occurs at a remarkably consistent age, despite varying genetic, environmental and lifestyle conditions.
Sophia modeled follicle depletion as a multi-stage stochastic process, similar to chemical reactions, where follicles transition through stages or die at specific rates. “We looked at the real data and found that the distribution of these points looks like a sequence of chemical reactions, so that is how we built this model — as a sequence of chemical reactions from one state to another state with a rate,” she explained.
Her simulations revealed that when the transition rates between different stages become comparable, the timing of menopause becomes synchronized across individuals, producing the narrow age distribution seen in real data. This framework not only accounts for the observed regularity of menopause but also suggests new ways to think about reproductive aging.
Looking back, Sophia credits her success to Kolomeisky’s mentorship. “He is an amazing person who has guided me through so many things. He has taught me a lot and makes the problems more interesting because he explains things behind the problem,” she said. With his guidance, she has built a strong foundation in both chemistry and statistics, and she plans to continue exploring this intersection in graduate school.