Boolean implications in transcriptomics
Başlık çevirisi mevcut değil.
- Tez No: 403329
- Danışmanlar: Prof. Dr. PETER F. STADLER, Dr. ANDREW HARRISON
- Tez Türü: Doktora
- Konular: Matematik, Mathematics
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2017
- Dil: İngilizce
- Üniversite: Universität Leipzig
- Enstitü: Yurtdışı Enstitü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 164
Özet
Özet yok.
Özet (Çeviri)
Background: Molecular phenotypes of cells are largely governed by their transcriptional state, i.e. by the amount and diversity of functional (protein coding mRNA and also noncoding ncRNA) RNA. With the invention of high throughput technologies such as microarrays and next generation sequencing it is possible to monitor thousands of RNA simultaneously, which in the end brings the possibility to estimate the gene expression state of the cells directly. Comparison of the expression levels between the genes enables to infer mutual relationships between them to study mechanisms of gene regulation often referred to as functional genomics. Analytical efforts have mostly concentrated on correlation analysis based on the assumption of co-expression of co-regulated genes and thus of a concerted change of the expression levels of the involved genes. However focusing solely on the co-expression restricts the search space to linear relations and naturally results in undirected networks. In contrast, Boolean implication analysis stands out as a straightforward yet robust and promising method to capture linear and non-linear pairwise relations between genes as well. It considers gene expression changes in Boolean space and is proven in modeling gene regulation and biological networks. Focus and Aim: In this thesis I provide a concept based on Boolean implication analysis that encapsulates; (i) identification of Boolean implications from the data, (ii) carrying on to network level analysis using these logical relations between genes, (iii) application to a 𝑀𝑌𝐶 network and casting regulatory relations to signals of biological functions as an example and, (iv) taking on to a systems biology approach by simulating the state space of the system. These methods were applied in combination with a battery of other supplementary methods and uses data of a microarray based cohort study on Germinal Centre - derived B-cell lymphoma. A software solution for application of Boolean implications to high throughput data is provided. The central objective of this thesis is development of a method of Boolean implications to analyze gene expression data and its evaluation in a proof-of-principle application to MYC-regulation in in lymphoma. My approach includes (down-top) data analysis and a reverse engineering (top-down) systems-biological modeling as well. Results: We used a self organizing maps pipeline (oposSOM) to organize, pre-process the data and to reveal functional gene modules that act in concerted/contrary fashion. Based on these preprocessed data we develop and examine in detail the Boolean implications method on levels supplied by SOM (gene, metagene and spot levels). A Boolean implication is simply a logical if/then relation between genes' expression. The method dichotomizes the expression range of genes into two regions (High Expression, Low Expression), and then, by comparing dichotomized expression values of gene couples, infers logical relations between them. The occurrence and general structure of these logical relations are supported by a theoretical approach by assuming expression profiles that can be described by Hill-equation. Next, we constructed a network where genes are nodes and implications are edges. We used an algorithm called Pathway Signal Flow (PSF) to calculate signals that represent functional biological output of the network and compared these signals to signals calculated for a MYC-network introduced by experimental literature in Burkitt's Lymphoma biology. We detected disturbed interactions that deregulate MYCfunction in a Lymphoma-subtype specific way and thus facilitate tumor progression. Finally, we interpreted and employed Boolean implications between genes as state updating rules and we explored the gene expression state space via a dynamical modeling approach. It provides possible progression paths of disease progression in its epigenetic (Waddington-) landscape and enables discovery of regulatory modes governing the formation of a particular disease subtype. Summary and Conclusion: As a summary we present a method of analyzing gene expression data using Boolean implications as an alternative to methods based on correlation analysis. Disadvantages of implication analysis is that it is computationally more expensive compared with correlation based methods and that it lacks a single parameter like correlation coefficient that describes the relation between two entities. On the other hand Boolean implication analysis equips one with information that enables the analysis of networks with higher resolution and thus to detect subtle regulatory modes and to assess their character in terms of six possible implications. The method is applicable to other high throughput data types such as DNA-methylation (high versus low methylation levels of the genes), mutations (mutated versus wild type), histone-modifications (modified versus unmodified), protein expression and metabolite abundance and combinations of them. My thesis established the methodical basis for such amendments with high impact for the discovery of mechanisms of genomic regulation.
Benzer Tezler
- State of emergency and international alignment behaviors: Origins and implications
Olağanüstü hâl ve uluslararası iş birliği davranışları: Kökenleri ve yansımaları
EMİNE ARI
Doktora
İngilizce
2022
Uluslararası İlişkilerKoç ÜniversitesiSiyaset Bilimi ve Uluslararası İlişkiler Ana Bilim Dalı
DOÇ. DR. REŞAT BAYER
- Sahada programlanabilir kapı dizileri ile lojik devre tasarımı ve VHDL kullanılarak bazı devrelerin gerçekleştirilmesi
Başlık çevirisi yok
ATEŞ BERNA
Yüksek Lisans
Türkçe
1998
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. AHMET DERVİŞOĞLU
- Programlanabilir lojik kontrolörler için tasarım yöntemleri
Başlık çevirisi yok
ERDAL ÖZKAN
Yüksek Lisans
Türkçe
1996
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrolİstanbul Teknik ÜniversitesiDOÇ.DR. SALMAN KURTULAN
- Runtime analysis of evolutionary algorithms with complex fitness evaluation mechanisms
Başlık çevirisi yok
DOGAN CORUS
- Zeroability patterns of monomials in the sign-representation of boolean functions
Boolean fonksiyonların işaret gösterimindeki terimlerin sıfırlanma düzenleri
OYTUN YAPAR
Yüksek Lisans
İngilizce
2017
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolÖzyeğin ÜniversitesiBilgisayar Bilimleri Ana Bilim Dalı
DOÇ. DR. ERHAN ÖZTOP