Integration and Analysis of Heterogeneous Biological Data

We live in the era of networks. The power of networks is the most fundamental driving force behind the machinery of life. Living bodies stay alive through complex inter-regulations of biochemical networks and information flows through these networks with such a great intensity and complexity that it exceeds anything that the human ingenuity has been able to spawn so far. Due to this overwhelming complexity we have begun to see a rapid rise in studies aimed at explaining the fundamental concepts and hidden properties of such complex systems. This thesis provides a strong foundation of using networks to understand complex biological phenomenon like protein functions, as well as more accurate method of modeling gene regulatory networks. In the first part we presented a methodology that uses existing biological data with gene ontology functional dependencies to infer functions of uncharacterized proteins. We combined different sources of structural and functional information along with gene ontology based term-specific relationships to predict precise functions of unannotated proteins. Such term-specific relationships, defined to clearly identify the functional contexts of each activity among the interacting proteins, which enables a dramatical improvement of the annotation accuracy with respect to previous approaches. The presented methodology may be easily extended to integrate more sources of biological information to further improve the function prediction confidence. In the second part of this thesis we discussed an extended BN model to account for post-transcriptional regulation in GRN simulation. Thanks to this extended model, we discussed the set of attractors of two biologically confirmed networks, focusing on the regulatory role of miR-7. Attractors have been compared with networks in which the miRNA was removed. The central role of the miRNA for increasing the network stability has been highlighted in both the networks, confirming the cooperative stabilizing role of miR-7. The enhanced BN model presented in this thesis is only a first step towards a more realistic analysis of the high-level functional and topological characteristics of GRNs. Resorting to the tool facilities, the dynamics of real networks can be analyzed. Thanks to the extended model that includes post-transcriptional regulations, not only the network simulation can be more reliable, but also it can offer new insights on the role of miRNAs from a functional perspective, and this improves the current state-of-the-art, which mostly focuses on high-level gene/gene or gene/protein interactions, neglecting post-transcriptional regulations. Due to its discrete nature, the BN model may still neglect some regulatory fine adjustments. However, the largest number of the computed attractors, now including miRNAs, still represents meaningful states of the network. The simple glimpse into the complexity of the network dynamics, that the toolkit is able to provide, could be used not only as a validation of in vitro experiments, but as a real System Biology tool able to rise new questions and drive new experiments.