Browse Theses

Despite increasing knowledge about pathophysiological pathways and cellular processes involved in diseases, the molecular mechanisms and physiological significance are not fully understood. Consequently, within this exploratory research we wish to lay the foundations for developing bioinformatics tools and systems biology approaches towards the analysis and modeling of transcriptional dynamics and the understanding of gene transcriptional regulatory program. Two in vivo models, namely corticosteroid pharmacogenomics in rat and human endotoxemia in human, have been investigated to gain insights into (1) adverse-effects, tissue-specificity, and circadian effects under corticosteroid treatment, (2) temporal regulatory programs in acute inflammation, and (3) cellular variability and synchronization as well as time-dependent systemic responses under acute stress. In order to pursue these goals, the hypothesis that informative components of the genome-wide transcriptional dynamics are composed of genes which are either co-expressed and co-functional or co-expressed across multiple conditions has been pursued to identify significant genome-wide transcriptional signatures. Concepts from consensus clustering and meta-analysis have been explored to avoid the bias and assumption of each single clustering method/metric and handle challenges in the analysis of microarray data from heterogeneous sources. Subsequently, the mysteries and complexities of transcriptional regulation have been explored by using two main strategies, namely phylogenetic foot-printing and context-specific CRM search, to identify relevant transcriptional regulators and examine the putative temporal transcriptional regulatory program. Additionally, an in silico multi-level agent-based model of human endotoxemia has been constructed to gain insights into cellular behaviors and circadian effects under acute stress. The model captures stochastic transcriptional dynamics and critical aspects of the in vivo physiological human endotoxemia model. By defining novel hypothetical quantities such as the variability-based fitness and the synchronization level, we provided a step forward to the exploration of cell-to-cell variability and stochastic dynamics of cellular behaviors as well as predictive implications inferred from cellular variability. In summary, our work aims at (i) identification of critical transcriptional signatures and regulatory controls to provide a better understanding of system behaviors and (ii) simulation to understand the cellular behaviors and circadian effects within specific contexts.