Our paper on ridge estimation of network models from time-course omics data appeared online in the Biometrical Journal. It extends previous work on the penalized estimation of the first order vector autoregressive model, which captures temporal and contemporaneous relations among the variates of a system (e.g. pathway). The new work extends this to allow for second order time effects, a group factor, or time-varying covariates. Application to in vitro cervical cancer data yields an attanuated picture HPV-induced cellular transformation. Read more.