That a parasite would kill its hosts is a major puzzle in evolutionary biology. Indeed, in many cases, the parasite's fitness at the epidemiological level would be higher if the host would remain alive longer.
Why virulence? We work on hypotheses that explain why virulent strains would even persist. In short, these are the following:
How does virulence vary? Assuming that a non-zero level of virulence is optimal, several factors can affect this optimal level. Importantly, the qualitative effect of these factor can depend on the reason that maintains virulence in the first place (e.g. transmission-virulence trade-off can lead to different results than multiple infections). This is why we only list typical effects rather than the effect they have. Amongst the effects we study there are spatial structure (how individuals are connected), nature of within-host interactions (in the case of multiple infections), transmission mode (vertical vs. horizontal or direct vs. vector-borne) or heterogeneity amongst hosts (e.g. males and females).
Applications? We try to apply these general theories to specific cases such as HIV, the 1918-1919 influenza or ebola virus. One of the ways to proceed, which we argue can help join theory and data, it to adopt a fitness based approach that is to express parasite fitness at the epidemiological level and then determine whether strains with different virulences vary in fitness.