The inherent complexity and non-linearity of biological regulatory and signaling networks prevent us from manually discerning predictive comprehensive models from functional experiments and clinical data. For example, knocking down certain combinations of genes in amputated trunk pieces of planarian worms can result in the regeneration of a complete organism with cryptic double-head or double-tail phenotypes; likewise, depending on which sub-clonal cancer cells are pharmacologically targeted in a tumor, it can collapse and disappear or grow and proliferate even faster. Despite the availability of these functional datasets, we still lack mechanistic explanations of gene regulation, cell-cell signaling, and spatial cellular behaviors responsible for these phenotypes. To bridge this gulf separating functional datasets from an understanding of cellular behaviors and phenotypic outcomes, we take a systems biology approach to automate the discovery of genetic, signaling, and metabolic dynamic networks directly from formalized phenotypic experimental data and predict the outcomes of novel perturbations and treatments. Our computational approach includes novel formal ontologies and databases of surgical, genetic, and pharmacological experiments together with machine learning algorithms based on evolutionary computation that can directly mine these data to reverse-engineer de novo precise mathematical dynamic models. We demonstrated this methodology by automatically inferring the first comprehensive genetic regulatory model of planarian regeneration and a stochastic signaling network of melanoma phenotypes. We correctly predicted novel genetic regulatory elements and their resulting phenotypes when perturbed and, importantly, novel never-seen-before phenotypic outcomes and the precise pharmacological manipulations to obtain them. We are now paving the way for understanding the regulation (and dis-regulation) of forms and shapes in development, cell-cell interactions in cancer, and genetic-metabolic networks in synthetic biology.