Conventionally, biological signal transduction networks are analysed using experimental and theoretical methods to describe specific protein components, interactions, and biochemical processes and to model network behavior under various conditions. While these studies provide crucial information on specific networks, this information is not easily converted to a broader understanding of signal transduction systems. Here, using a specific model of protein interaction we analyse small network topologies to understand their response and general properties. In particular, we catalogue the response for all possible topologies of a given network size to generate a response distribution, analyse the effects of specific biochemical processes on this distribution, and analyse the robustness and diversity of responses with respect to internal fluctuations or mutations in the network. The results show that even three- and four-protein networks are capable of creating diverse and biologically relevant responses, that the distribution of response types changes drastically as a function of biochemical processes at protein level, and that certain topologies strongly pre-dispose a specific response type while others allow for diverse types of responses. This study sheds light on the response types and properties that could be expected from signal transduction networks, provides possible explanations for the role of certain biochemical processes in signal transduction and suggests novel approaches to interfere with signaling pathways at the molecular level. Furthermore it shows that network topology plays a key role on determining response type and properties and that proper representation of network topology is crucial to discover and understand so-called building blocks of large networks.