<p>This dissertation proposes a new working model of reductionism for biology and a new concept of the gene based on the new reduction model. My project aims to help biologists and philosophers understand what reductionism in biology really is, or, should be. Historical debates about reductionism testify us that the classical reduction model, i.e., Ernest Nagel's bridge-law model, offers us neither an appropriate ontological reductionism nor a reductive explanation about biological phenomena. Casting doubts on the received view of the layered hierarchical model of ontology, I suggest that many interesting biological properties be construed as second-order functional properties and their first-order realizers. Providing for reduction finely-analyzed biological properties, I offer a new model for reductionism in biology - localized functional reductionism - which evolved from Jaegwon Kim's view of reductionism presented for the problems of mental causation. </p><p>My localized functional reductionism shows that a localized functional property is reduced to its base/structural property. I emphasize that researchers in biology do not deal with abstract general properties but always localized, structure-specific biological properties. A localized functional property and the structure-specific biological property as its base property are what we are interested in and this is what makes biological properties appropriate for research and meaningful for philosophical discussion. The localized functional reduction model, which is actually a case of token reduction model, integrates the fine-grained ontological hierarchies of both macro/micro-levels and higher/lower-orders, and it also synthesizes functional reductionism and token identity thesis. In my localized functional reductionism, functional biological properties are not eliminated but they exist with their own causal powers and true explanatory powers. </p><p>I also argue that the gene, construed as a second-order functional property, must be understood as gene expression network-specific. The gene, when it is realized on a given occasion, is reduced to, and is identical with, one of its genomic realizers on the given occasion, that is, the gene expression network. A new dynamic approach to the concept of the gene as the gene expression network vindicates reductionism.</p> / Dissertation

Cancer presents as a highly heterogeneous disease with partly overlapping and partly distinct (epi)genetic characteristics. These characteristics determine inherent and acquired resistance, which need to be overcome for improving patient survival. In line with the global efforts in identifying druggable resistance factors, extensive preclinical research of the Cordes lab and others designated the cancer adhesome as a critical and general therapy resistance mechanism with multiple druggable cancer targets. In our study, we addressed pancancer cell adhesion mechanisms by connecting the preclinical datasets generated in the Cordes lab with publicly available transcriptomic and patient survival data. We identified similarly changed differentially expressed genes (scDEGs) in nine cancers and their corresponding cell models relative to normal tissues. Those scDEGs interconnected with 212 molecular targets from Cordes lab datasets generated during two decades of research on adhesome and radiobiology. Intriguingly, integrative analysis of adhesion associated scDEGs, TCGA patient survival and protein-protein network reconstruction revealed a set of overexpressed genes adversely affecting overall cancer patient survival and specifically the survival in radiotherapy-treated cohorts. This pancancer gene set includes key integrins (e.g. ITGA6, ITGB1, ITGB4) and their interconnectors (e.g. SPP1, TGFBI), affirming their critical role in the cancer adhesion resistome. In summary, this meta-analysis demonstrates the importance of the adhesome in general, and integrins together with their interconnectors in particular, as potentially conserved determinants and therapeutic targets in cancer.

Motivation: Disease progression in cancer can vary substantially between patients. Yet, patients often receive the same treatment. Recently, there has been much work on predicting disease progression and patient outcome variables from gene expression in order to personalize treatment options. A widely used approach is high-throughput experiments that aim to explore predictive signature genes which would provide identification of clinical outcome of diseases. Microarray data analysis helps to reveal underlying biological mechanisms of tumor progression, metastasis, and drug-resistance in cancer studies. Despite first diagnostic kits in the market, there are open problems such as the choice of random gene signatures or noisy expression data. The experimental or computational noise in data and limited tissue samples collected from patients might furthermore reduce the predictive power and biological interpretability of such signature genes. Nevertheless, signature genes predicted by different studies generally represent poor similarity; even for the same type of cancer. Integration of network information with gene expression data could provide more efficient signatures for outcome prediction in cancer studies. One approach to deal with these problems employs gene-gene relationships and ranks genes using the random surfer model of Google's PageRank algorithm. Unfortunately, the majority of published network-based approaches solely tested their methods on a small amount of datasets, questioning the general applicability of network-based methods for outcome prediction. Methods: In this thesis, I provide a comprehensive and systematically evaluation of a network-based outcome prediction approach -- NetRank - a PageRank derivative -- applied on several types of gene expression cancer data and four different types of networks. The algorithm identifies a signature gene set for a specific cancer type by incorporating gene network information with given expression data. To assess the performance of NetRank, I created a benchmark dataset collection comprising 25 cancer outcome prediction datasets from literature and one in-house dataset. Results: NetRank performs significantly better than classical methods such as foldchange or t-test as it improves the prediction performance in average for 7%. Besides, we are approaching the accuracy level of the authors' signatures by applying a relatively unbiased but fully automated process for biomarker discovery. Despite an order of magnitude difference in network size, a regulatory, a protein-protein interaction and two predicted networks perform equally well. Signatures as published by the authors and the signatures generated with classical methods do not overlap -- not even for the same cancer type -- whereas the network-based signatures strongly overlap. I analyze and discuss these overlapping genes in terms of the Hallmarks of cancer and in particular single out six transcription factors and seven proteins and discuss their specific role in cancer progression. Furthermore several tests are conducted for the identification of a Universal Cancer Signature. No Universal Cancer Signature could be identified so far, but a cancer-specific combination of general master regulators with specific cancer genes could be discovered that achieves the best results for all cancer types. As NetRank offers a great value for cancer outcome prediction, first steps for a secure usage of NetRank in a public cloud are described. Conclusion: Experimental evaluation of network-based methods on a gene expression benchmark dataset suggests that these methods are especially suited for outcome prediction as they overcome the problems of random gene signatures and noisy expression data. Through the combination of network information with gene expression data, network-based methods identify highly similar signatures over all cancer types, in contrast to classical methods that fail to identify highly common gene sets across the same cancer types. In general allows the integration of additional information in gene expression analysis the identification of more reliable, accurate and reproducible biomarkers and provides a deeper understanding of processes occurring in cancer development and progression.:1 Definition of Open Problems 2 Introduction 2.1 Problems in cancer outcome prediction 2.2 Network-based cancer outcome prediction 2.3 Universal Cancer Signature 3 Methods 3.1 NetRank algorithm 3.2 Preprocessing and filtering of the microarray data 3.3 Accuracy 3.4 Signature similarity 3.5 Classical approaches 3.6 Random signatures 3.7 Networks 3.8 Direct neighbor method 3.9 Dataset extraction 4 Performance of NetRank 4.1 Benchmark dataset for evaluation 4.2 The influence of NetRank parameters 4.3 Evaluation of NetRank 4.4 General findings 4.5 Computational complexity of NetRank 4.6 Discussion 5 Universal Cancer Signature 5.1 Signature overlap – a sign for Universal Cancer Signature 5.2 NetRank genes are highly connected and confirmed in literature 5.3 Hallmarks of Cancer 5.4 Testing possible Universal Cancer Signatures 5.5 Conclusion 6 Cloud-based Biomarker Discovery 6.1 Introduction to secure Cloud computing 6.2 Cancer outcome prediction 6.3 Security analysis 6.4 Conclusion 7 Contributions and Conclusions