The core message of this work is to develop methods for incorporating expert or background knowledge about causal relationships into the analysis of maximal ancestral graphs (MAGs) in order to restrict the Markov equivalence class and improve causal identification.
