We develop and apply techniques for inferring topologies, causal structures and parameters of biological networks from wet lab data. We also develop sensitivity analysis methods to assess the impact of changing both variables and parameters in the model of inferred networks. We identify gene and protein pathways linked to diseases onset and progression and clarify models of mechanisms of action of drugs as well as predicting networks whose structure and dynamics are linked to the healthy status or to pharmacological and/or dietary intervention. We also exploit methods in ecology to predict structures of interaction among individuals as well as vulnerabilities of ecosystems.
We develop and apply network analysis for identifying key genes representing potential gatekeepers linking nutrition to disease. Network communities significantly enriched in proteins responding to drug pharmacology can highlight unexpected mechanisms of functioning, beyond the boundaries of canonical metabolic and regulatory pathways. Similarly, network analysis can shed light on the forces driving the evolution of food webs, focusing on the connection between ecosystem structure and dynamics.
We develop and apply simulation-based techniques that start from the list of components of the system, their (partial) wiring and their partial quantitative characterization. Signaling and metabolic networks are the main networks we model, simulate and analyze, linked to molecular nutrition and systems pharmacology. Once validated, models can be used as predictive tools by perturbing them with addition/deletion/modification of components.