The pathogenicity of L. pneumophila, the causative agent of Legionnaires’ disease, depends on an arsenal of interacting proteins. Here we describe how surface-associated and secreted virulence factors of this pathogen interact with each other or target extra- and intracellular host proteins resulting in host cell manipulation and tissue colonization. Since progress of computational methods like AlphaFold, molecular dynamics simulation, and docking allows to predict, analyze and evaluate experimental proteomic and interactomic data, we describe how the combination of these approaches generated new insights into the multifaceted “protein sociology” of the zinc metalloprotease ProA and the peptidyl-prolyl cis/trans isomerase Mip (macrophage infectivity potentiator). Both virulence factors of L. pneumophila interact with numerous proteins including bacterial flagellin (FlaA) and host collagen, and play important roles in virulence regulation, host tissue degradation and immune evasion. The recent progress in protein-ligand analyses of virulence factors suggests that machine learning will also have a beneficial impact in early stages of drug discovery.