Universität zu Köln

Leote, Ana Carolina ORCID: 0000-0003-0879-328X (2023). Interpreting gene expression changes in single cells and during ageing using transcriptional regulatory networks. PhD thesis, Universität zu Köln.

PDF (Doctoral Dissertation) DoctoralDissertation_AnaCarolinaLeote_published.pdf - Published Version Bereitstellung unter der CC-Lizenz: Creative Commons Attribution Non-commercial. Download (16MB)

Abstract

Regulation of gene expression allows for cells to execute highly specialized functions and orchestrate a variety of responses to stimuli. Impaired coordination of gene expression is linked to several diseases – including ageing – highlighting the relevance of comprehensively understanding gene expression coordination. However, current models of gene expression coordination are often limited to specific biological conditions or focused on individual cellular processes. Here, we report a transcriptional regulatory model, derived from more than 1000 expression datasets, able to capture tissue- and cell-type-specific gene-gene relationships, as well as global (cross-tissue) relationships. We take advantage of the wide applicability of this model to tackle two distinct biological problems. The first concerns the computational estimation of missing values in single-cell RNA-sequencing (scRNA-seq) data. scRNA-seq methods are typically unable to quantify the expression levels of all genes in a cell (dropout events). Several methods have been proposed for the estimation of dropout expression (dropout imputation), with no clear method outperforming others across datasets and downstream analyses. We propose a new method that makes use of the transcriptional regulatory model to estimate the expression of dropouts and show it outperforms published state-of-the-art methods, especially for lowly expressed genes, including cell-type-specific transcriptional regulators. We observed gene- and dataset-dependent performance of the methods we tested, leading us to implement an R package, ADImpute, that automatically determines the best imputation method for each gene in a dataset. This work represents a paradigm shift by demonstrating that there is no single best imputation method. Instead, we propose that imputation should maximally exploit external information and be adapted to gene-specific features, such as expression level and expression variation across cells. The second problem addressed in this work was the impact of ageing on gene expression coordination. We use existing RNA-seq data of human tissues at different ages to investigate the impact of ageing on the gene-gene relationships captured in the transcriptional regulatory model. We observed age-related changes towards both a strengthening and a loosening of gene-gene relationships with age, mostly impacting genes with mitochondrial functions and cell cycle regulation. We detected age-related changes in relationships between genes involved in the same functional module, as well as genes in distinct functional modules, highlighting the impact of ageing on the coordination of cellular processes. This work demonstrates the importance of zooming out of the effect of ageing on individual genes or cellular processes and investigating how their crosstalk is affected at a systems level. Put together, the work presented here shows that transcriptional gene-gene relationships can be ‘learned’ from a rather limited set of example datasets and subsequently applied to a wide range of cell- and tissue types, where pathological breakdown of gene-gene relationships can be investigated.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Leote, Ana Carolina anacarolinaleote@gmail.com orcid.org/0000-0003-0879-328X UNSPECIFIED
URN:	urn:nbn:de:hbz:38-716414
Date:	30 November 2023
Language:	English
Faculty:	Faculty of Mathematics and Natural Sciences
Divisions:	CECAD - Cluster of Excellence Cellular Stress Responses in Aging-Associated Diseases
Subjects:	Life sciences
Uncontrolled Keywords:	Keywords Language Gene expression regulation English Aging English Single-cell RNA-sequencing English
Date of oral exam:	23 November 2022
Referee:	Name Academic Title Beyer, Andreas Prof. Dr. Tessarz, Peter Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/71641