Inference error versus number of random valid edges provided. The proportional error (i.e., inference error) denotes the minimal cross-validation error divided by the minimal least-squares error of the linear regression without any regularization terms and averaged over five random networks and five random sets of valid edges. Zero to 25 (i.e., the number of all edges in each random network) prior edges are provided respectively. (A), (B) and (C) are the results of 4, 5 and 6 observations, respectively. The proportional errors decrease when more valid edges are provided. Prior connections appear to be more important when only a few observations are available.