Constructing and Analyzing Dynamic Biological Interaction Networks
Complex interactions among biological entities can be effectively modeled as interaction networks where nodes represent entities of interest such as proteins, genes or complexes and edges mimic associations among them. The construction and analysis of these interaction networks form an important step in annotating unknown proteins and genes, determining regulatory relations, identifying drug targets, and integrating diverse datasets. Recently, the size and complexity of biological datasets have undergone a rapid expansion due to advances in high-throughput screening techniques. Therefore, the need for sophisticated computational techniques to convert diverse and noisy biological datasets into interaction networks and to study these networks for useful information has increased. A further challenge lies in exploring dynamic interactions among cellular components, such as genes, proteins, and metabolites, from these datasets. In this talk, I will discuss our probabilistic model for integrating heterogeneous evidence to predict interactions between regulatory proteins and genes of the Budding Yeast organism. We showed that by integrating direct protein-DNA interactions from ChIP-chip measurements with transcription factor motif hits and nucleosome occupancy datasets, one can significantly improve the quality of predicted interactions. Moreover, our methodology enabled us to predict condition-specific interactions under six stress conditions as well as the normal growth conditions. We further investigated predicted interactions along with gene expression measurements in order to characterize their regulatory implications. Consequently we employed a Data Mining technique to determine potential factors that play a role in the regulatory behavior of protein-DNA interactions. Our analysis produced testable hypotheses on dynamic regulatory interactions and their causes.