Epidemic dynamics has been traditionally studied in homogeneously mixed and infinitely large populations. However, real populations are finite and highly heterogeneous. We aim to develop individual-based mathematical models to describe epidemic processes in populations that are characterised by high heterogeneity in the characteristics of individuals and complex interactions among them. The focus will be on the identification of the properties of the contact network structure that could predict the impact of the epidemics in different real-world scenarios.