2025 Theses Doctoral

Computational Results and Insights into Cancer-Associated Fibroblast Formation and Tumor Heterogeneity

Cai, Lingyi

This dissertation uncovers a pan-cancer cellular transition of a progenitor population of adipose stromal cells (ASCs) to “aggressive” cancer-associated fibroblasts (aCAFs), associated with cancer invasiveness, poor prognosis, metastasis, and chemoresistance.

Through comprehensive computational analyses of single-cell RNA sequencing (scRNA-seq) data across multiple cancer types and normal tissues, we trace this transition to a previously uncharacterized progenitor population with both adipogenic and fibroblastic differentiation potential, found naturally in cancer-free individuals as ASCs in adipose tissue and fibro-adipogenic progenitors (FAPs) in skeletal muscle.

We identified SFRP4 and LINC01614 as key regulators of this ASC/FAP to aCAF transition and validated their roles experimentally using CRISPR knockouts and in vitro two-dimensional co-culture experiments. Recognizing limitations in existing computational tools, we developed CASCC (co-expression-assisted single-cell clustering), a novel clustering method that improves biological accuracy in single-cell transcriptomic analyses.

Finally, we extended our study to three-dimensional patient-derived organoids, integrating transcriptomic and genomic data to better model tumor heterogeneity and reveal additional therapeutic targets. Altogether, this work improves our understanding of tumor microenvironments and provides new computational methods and strategies for biomolecular data.