In this work we consider structure discovery of undirected graphical models from observational data. Inferring likely structures from few examples is a complex task often requiring formulating priors and sophisticated inference procedures. In the setting of Gaussian Graphical Models (GGMs) a popular approach to formulating an estimator is with a penalized maximum likelihood objective on the precision matrix. This objective is often difficult to design to specifically fit ones priors and the graph structure recovery is often not explicitly possible to embed in the objective, moreover incorporating any additional assumptions often requires a great deal of research effort. By contrast, it may be easier to generate samples of data that are arise from graphs with the desired properties. We propose here to leverage this latter source of information in order to learn a function that maps from empirical covariance matrices to estimated graph structures. This learned function brings two benefits: it implicitly models the desired structure or sparsity properties to form suitable priors, and it can more directly be tailored to the specific problem of edge structure discovery. We apply this framework to several critical real world problems in structure discovery and show that it can be competitive to standard approaches such as graphical lasso, at a fraction of the execution speed. We use deep neural networks to parametrize our estimators. Experimentally, our learn able graph discovery method trained on synthetic data generalizes well to different data: identifying relevant edges in real data, completely unknown at training time. We find that on genetics, brain imaging, and simulation data we obtain competitive (and often superior) performance, compared with analytical methods.