Repository logo
 

Systems immunology frameworks link multicellular immune perturbation phenotypes and setpoints to response outcomes


Type

Thesis

Change log

Authors

Mulé, Matthew 

Abstract

This thesis develops frameworks for using top-down systems biology approaches with multiomic single cell technology to understand variation in human immune system response outcomes. We integrate human population, cell subset and single cell variations in molecular phenotypes which give rise to baseline setpoints, are perturbed by vaccination or drug treatment, and are linked to emergent response / clinical outcomes. In Chapter 2, we dissect noise sources in data derived from new methods which simultaneously measure protein and mRNA in single cells (e.g., CITE-seq). After identifying two main sources of noise, we develop an open source software method for normalizing and denoising CITE-seq protein data. We then develop a computational framework for analyzing the effects of timed immune system perturbations applied to human cohorts profiled with multimodal single cell technology. Chapter 3 applies these methods on a human vaccination cohort profiled using CITE-seq. We first define highly interpretable immune cell subsets using unsupervised clustering based on the denoised protein data, then contrast the transcriptome pathways within these subsets that are differentially induced by vaccines formulated with and without an adjuvant. These robust phenotypes are further interpreted using single cell computational reconstructions of cell states. Using these comparative analyses, along with unbiased analysis of baseline phenotypes linked to antibody response, we identify a multicellular “naturally adjuvanted” human immune system setpoint more poised to respond to vaccination. Chapter 4 applies the methods developed above to a cohort of cancer patients treated with immune checkpoint inhibitors. In this work, we identify multicellular baseline setpoints linked to development of immune related adverse events after treatment which are uncoupled from the phenotypes induced by treatment. Together, these approaches help advance a quantitative, predictive understanding of human immune system variation, and pave the way for further human perturbation cohort studies across biological disciplines.

Description

Date

2022-09-29

Advisors

Smith, Ken
Tsang, John

Keywords

systems immunology, immunology, single cell

Qualification

Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge