Development of computational methods for biological complexity

The cell is a complex system. In this system, the different layers of biological information establish complex links converging in the space of functions; processes and pathways talk each other defining cell types and organs. In the space of biological functions, this lead to a higher order of “emergence”, greater than the sum of the single parts, defining a biological entity a complex system. The introduction of omic techniques has made possible to investigate the complexity of each biological layer. With the different technologies we can have a near complete readout of the different biomolecules. However, it is only through data integration that we can let emerge and understand biological complexity. Given the complexity of the problem, we are far from having fully understood and developed exhaustive computational methods. Thus, this make urgent the exploration of biological complexity through the implementation of more powerful tools relying on new data and hypotheses. To this aim, Bioinformatics and Computational Biology play determinant roles. The present thesis describes computational methods aimed at deciphering biological complexity starting from genomic, interactomic, metabolomic and functional data. The first part describes NET-GE, a network-based gene enrichment tool aimed at extracting biological functions and processes of a set of gene/proteins related to a phenotype. NET-GE exploits the information stored in biological networks to better define the biological events occurring at gene/protein level. The first part describes also eDGAR, a database collecting and organizing gene-disease associations. The second part deals with metabolomics. I describe a new way to perform metabolite enrichment analysis: the metabolome is explored by exploiting the features of an interactome. The third part describes the methods and results obtained in the CAGI experiment, a community experiment aimed at assessing computational methods used to predict the impact of genomic variation on a phenotype.