Author ORCID Identifier

Date of Award

Spring 5-31-2023

Document Type

Thesis (Undergraduate)


Computer Science

First Advisor

Sarah M. Preum

Second Advisor

Joshua J. Levy


Skin photoaging is the premature aging of skin that results from ultraviolet light exposure. It is a major risk factor for the development of skin cancer, among other malignant skin pathologies. Accordingly, understanding its etiology is important for both preventative and reparative clinical action. In this study, skin samples obtained from patients with ranging solar elastosis grades – a proxy for skin photoaging – were sequenced using next-generation sequencing techniques to further understand the genomic, epigenomic, and histological signs and signals of skin photoaging. The results of this study suggest that tissues with severe photoaging exhibit increases in the frequency of some immune cell populations, especially CD4, CD8, and regulatory T cells, macrophages, and mast cells, and decreases in the frequency of other immune cell populations, like dendritic cells, NKT cells, and plasma cells. Samples with severe solar elastosis also had increased expression of genes associated with the complement cascade and broader innate immune system. Methods to infer spatial transcriptomics data using purely histomorphological data were also devised in this work. Vision Transformers were trained to infer spatially variable gene expression across dichotomized and continuous regression tasks, reaching a median average AUC of 0.80 on the dichotomized task and a median average Spearman coefficient of 0.60 on the continuous task. Finally, methods were devised to spatially resolve chromatin accessibility data across whole slide images, and models were trained to predict spatial chromatin accessibility, reaching a median AUC score of 0.71. These results suggest that deep learning can be used to democratize access to spatial transcriptomics insights, facilitating a deeper understanding of disease etiologies informed by cell and tissue spatial dynamics.