Analysis of cellular heterogeneity in breast cancer by single cell sequencing

Thesis

Breast cancer is a complex heterogenous 3D ecosystem. The heterogenous composition of breast cancer determines disease progression and treatment responses. Triple receptor negative breast cancer (TNBC) is a distinct subtype with poor clinical outcomes. Deconvolution of spatially-regulated transcriptomic and microenvironmental drivers unique to TNBC offers the potential to reveal new therapeutic vulnerabilities.

Single cell RNA sequencing (scRNA-seq) and spatial transcriptomic technologies were applied to three treatment naive patient-derived breast cancer samples. New spatial transcriptomic and scRNA-seq experimental pipelines were established. The new technologies were successfully applied to clinical grade biopsy samples. Cellular heterogeneity within the epithelial and non-epithelial compartment was identified across the three samples. The heterogeneity identified is consistent with the published literature.

Knowledge in the theoretical underpinnings for scRNA-seq analysis along with the skills required for data analysis in a small patient cohort were acquired during the DPhil. The application of algebraic topology, manifold learning and graph theory in evaluating and interpreting scRNA-seq has been studied.

The computational tools available for integrating spatial transcriptomics and scRNA-seq data were critically appraised. Future perspectives on approachesfor multimodal integration were explored.

Authors: Sayal, K

• DOI: 10.5287/ora-n1evy0myg
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: manifold theory; spatial transcriptomics; single cell sequencing; breast cancer; machine learning
• Subjects: Translational Oncology; Machine Learning