The advantage provided by the increasingly interconnected nature of our world has generated a dangerous by-product: the possibility for rapid worldwide spread of epidemics. The ability to forecast epidemic evolution – as much accurately as we can now do for weather conditions – would be of invaluable help in fighting the emergence or re-emergence of viruses such as SARS, avian influenza, HIV-AIDS, Lyme disease, West Nile virus, or more recently the threat of an influenza pandemic. With the advent of interdisciplinary tools and methods, the latest modeling approaches for the study of the spread and control of infectious diseases witness the emergence of a new area of research – computational epidemiology – that integrates mathematical and statistical epidemiology with computational sciences and informatics tools to conduct scenario analysis in public health domain. While few research groups have begun to use large scale simulations for epidemic modeling, many fundamental theoretical questions are left unanswered. How does the complex nature of real world affect our predictive capabilities in the realm of computational epidemiology? What are the fundamental limits in epidemic evolution predictability with computational modeling? How do they depend on the level of accuracy of our description and knowledge of the state of the system? The present project aims at developing a vigorous research effort along two main directions corresponding to i) the formulation of models for the basic theoretical understanding of multi-scale and agent based approaches and their predictive power; ii) the development of computational approaches and data integration tools that will provide a realistic modeling framework for the analysis of observed epidemic outbreaks and the forecast of patterns of emerging diseases. The ERC Starting Independent Researcher Grant offers an ideal opportunity to start a structured program in this direction, aimed at providing fundamental advances in the field.