title: Dimensionality reduction methods for high-dimensional biological data analysis
creator: Capraz, Klemens Tümay
description: Disease progression and response to treatments can strongly differ between  patients, due to each individual’s unique genetic, environmental and molecular factors. Precision medicine aims to understand these factors in disease context and tailor treatments to patients based on their molecular pro-  file. High-throughput multi-omics techniques are powerful tools to measure  molecular profiles of cells, tissues and organs. They offer exciting possibilities in precision medicine and biomarker discovery, because they allow for  comprehensive and large-scale measurements of genes, proteins, metabolites, microbiome composition and other biological information. The gained  data’s complexity and high-dimensionality provide an opportunity to capture a lot of useful information about a biological sample. However, the  vastness of the data also makes interpretation challenging. Moreover, we  often do not have access to a large number of samples in biological settings,  which poses additional challenges to data analysis. To be able to understand  and identify patterns in complex multi-omics data, we need dimensionality  reduction methods. In this thesis, I start with presenting a short overview  of the omics modalities we can measure, I explain how they can be used in  precision medicine, and I present the most common dimensionality reduction methods used in biological data analysis. In the subsequent chapters I  introduce novel methods I developed and used for high-throughput biological data analysis. I present a new feature selection method based on replicate reproducibility, a Factor Analysis model that allows the incorporation of  prior knowledge, and a computational method to generate mutant libraries  for high-throughput screening for reverse genetics studies of gut microbes.  All methods I introduce are implemented in Python or R packages and are  available as open-source software.
date: 2024
type: Dissertation
type: info:eu-repo/semantics/doctoralThesis
type: NonPeerReviewed
format: application/pdf
identifier: https://archiv.ub.uni-heidelberg.de/volltextserver/35515/1/thesis_capraz.pdf
identifier: DOI:10.11588/heidok.00035515
identifier: urn:nbn:de:bsz:16-heidok-355153
identifier:   Capraz, Klemens Tümay  (2024) Dimensionality reduction methods for high-dimensional biological data analysis.  [Dissertation]     
relation: https://archiv.ub.uni-heidelberg.de/volltextserver/35515/
rights: info:eu-repo/semantics/openAccess
rights: Please see front page of the work (Sorry, Dublin Core plugin does not recognise license id)
language: eng