Life is a self-sustaining and self-regulating chemical reaction network. In other words, a living system continuously regenerates its own components, in such a way that these components maintain and regulate the underlying reaction network that produced them. Autocatalytic sets are a formalization of this notion of life. An autocatalytic set is a reaction network in which each reaction is catalyzed by at least one of the molecules from the set itself, and each molecule can be produced from a suitable food source by using only reactions from the set itself. Autocatalytic sets are believed to have played an important role in the origin and early evolution of life. My colleague Prof. Mike Steel and I have developed a formal framework known as RAF theory to detect and analyze autocatalytic sets in general reaction networks. We have shown that autocatalytic sets are highly likely to exist under a wide variety of realistic assumptions in simple computational models of chemical reaction networks, and that these sets usually have a rich hierarchical structure of smaller and smaller autocatalytic subsets, which is an important requirement for their possible evolution. Furthermore, we have shown that the formal framework can be successfully applied to real chemical and biological networks as well, and that autocatalytic sets indeed exist in such real networks. However, there is still somewhat of a gap between theory and experiments on the emergence and further evolution of autocatalytic sets. In this project, I will continue and expand my research on autocatalytic sets, in particular in collaboration with experimental chemists, in an effort to close this gap.