This communication presents a mathematical mechanism-based model of the regenerating liver after drug-induced pericentral liver lobule damage that implements alternative hypotheses on the mechanistic interplay of cell types to permit predictions of cell perturbation experiments. Start point is a minimal spatial temporal model of liver regeneration omitting all other cell types than hepatocytes and blood vessels. These are hepatic stellate cells, Kupffer cells and infiltrating macrophages, neutrophils, and platelets. Single-cell (agent)-based models were used for simulation of multicellular tissue organization at the level of liver tissue microarchitecture where each agent is parameterized by biophysical and biokinetic parameters permitting determination of meaningful parameter ranges. The model is set up in a virtual liver architecture and integrates all relevant cell types and intercellular signals. The interplay between the systems components reflects current hypotheses and mechanisms. Results: A model explaining the experimentally found scenarios of liver regeneration as well as published knowledge that includes the relevant cell types was established. In a first step, the consequence of alternative hypotheses about the interplay of different cell types on regeneration have been studied. Regeneration has been quantified by the size of the dead cell area, the hepatocyte density and the explicitly spatial-temporal profile of all cell types. If for example infiltrating macrophages do not contribute to elimination of dead hepatocytes, or if TGFβ activation of hepatic stellate cells (HSCs) is lacking, the lesion is not closed in time. To demonstrate how this model can be used to guide experimental design, it has then been used to predict the outcome of perturbation experiments. If for example HSCs are knocked down, the initial dead cell area appears smaller but is not closed in time, while if infiltrating macrophages are knocked down, the model predicts incomplete regeneration. Deviation of the observed from the simulated system trajectories indicate branching points at which the systems behavior cannot be explained by the underlying set of hypotheses anymore and hence require further experimental investigations. The outcome points towards a successful strategy towards a full digital liver twin that permits to test perturbations from the molecular up to the cell, tissue, and body scale. It is complementing the experiment by guiding towards the most informative experiments, identifying gaps in mechanistic knowledge and those experimental designs that are expected to generate an informative outcome. Such a systematic strategy ultimately permits to focus and economize resources and can prospectively be speeding up the process of mechanism identification. It is also able to guide diagnosis and therapy if fed with patient data. In so far, our work responds on the question of systems complexity.