East Boston High School
Many cancer cells acquire stem cell-like phenotypes, which are marked by dedifferentiation away from a mature cell state. For these types of cancers, differentiation therapy is an attractive alternative to traditional chemotherapy, which indiscriminately targets both cancer and healthy cells and has severe side effects for patients. In differentiation therapy, various drugs affect gene expression so that cells differentiate into more "normal" cell types. This does not kill cells, but can have the effect of decreasing the malignancy of the cancer. Using statistical techniques such as principal component analysis, and non-parametric methods such as the Wilcoxon Rank-Sum test and z-scores, Stephanie analyzed gene expression data from Connectivity Map (CMAP) cell lines treated with various small molecule drug candidates to determine which drugs had the largest effect on genes related to development. Using this dataset, Stephanie identified 12 drug candidates that may be useful in future differentiation therapy studies.
Stephanie has long been interested in applying mathematics and computer science to real-world problems. “[My mentor] taught me many things from programming techniques and how one must be meticulous with conducting scientific research. There are downsides to certain statistical tests, and some conclusions may not be definitive,” said Stephanie. “The entire process of conducting the statistical analysis for my project made me realize that I truly enjoy utilizing mathematics in science.”