2026-09-23T11:02:37Zhttps://api.repository.cam.ac.uk/server/oai/request

oai:www.repository.cam.ac.uk:1810/3917752025-12-20T01:46:49Zcom_1810_263977com_1810_221767com_1810_256067col_1810_263989

Decoding Human T Cell Immunity With Single-Cell Multi-Omics https://doi.org/10.17863/CAM.122780 Dratva, Lisa Marietta 0000000228736787 Teichmann, Sarah In this work, I explore human adaptive immunity through single-cell omics by jointly analysing T cell gene expression and T cell receptor (TCR) sequence modalities. Across several longitudinal infectious disease studies, I identify transient T cell states after infection and link these phenotypes with specific TCR sequences. In the first chapter, I provide a comprehensive overview of the field's state-of-the-art, encompassing experimental single T cell profiling methods, relating key findings from prior single-cell studies, and computational tools for antigen specificity prediction. Throughout the next chapters, I introduce Cell2TCR, an open-source framework for cross-donor TCR sequence comparison, and make the case for convergent clonal selection after infection through the discovery and characterisation of TCR motifs. Next, I link TCR motifs to antigen specificity using experimental T cell specificity databases and validate the findings with in-house data in the context of several viral infections, by integration with public single-cell data, and by comparing to bulk sequencing samples. I demonstrate the framework's utility for motif discovery in a controlled human infection study as well as in the context of community infections within an at-risk cohort, and make it compatible with human gamma-delta T cell and murine T cell responses. Furthermore, I show that TCR motifs exhibit T cell compartment preference and strong donor major histocompatibility complex (MHC) restriction, consistent with TCR-peptide-MHC recognition principles. In the last results chapter, I present a T cell atlas comprising 3,6 million T cell transcriptomes, and with paired TCR sequences for about half of them, which has been curated from 23 datasets of various infectious diseases. I outline a strategy to propose novel peptide-MHC combinations for a given disease context like infection, integrate 3D structure-informed model ranking into the process and evaluate it using internally generated antigen specificity data. Taken together, this research represents a significant step towards de novo prediction of T cell antigen specificity, paving the way for large-scale TCR repertoire interpretation and deliberate TCR engineering. University of Cambridge 2025-06-11 Thesis Doctoral Doctor of Philosophy (PhD) eng https://www.repository.cam.ac.uk/handle/1810/391775 https://www.repository.cam.ac.uk/bitstreams/0042efaa-db77-4c98-83bc-55f556f2cbbb/download e1868a2b18da321611123def3d017151 https://www.repository.cam.ac.uk/bitstreams/bfaa24e5-f0d3-4f54-9e32-7c94a485327d/download 87eda9de84448d1f82354d60eee3eb5f http://purl.org/NET/rdflicense/allrightsreserved Immunology Single-cell T cell TCR Antigen specificity Genomics