154 Bayesian Networks: With Examples in R
The BNs learned with these two approaches are shown in Figure 6.6. Some
of the structural features detected in Sachs et al. (2005) are present in both
dag.wh and dag.tiers. For example, the interplay between Plcg, PIP2 and
PIP3 and between PKC, P38 and Jnk are both modelled correctly. The lack of
any direct intervention on PIP2 is also correctly modelled in dag.tiers. The
most noticeable feature missing from both DAGs is the pathway linking Raf
to Akt through Mek and Erk.
The approach used in Sachs et al. (2005) yields much better results. In-
stead of including the interventions in the network as an additional node,
they used a modified BDe score (labelled "mbde" in bnlearn) incorporating
the effects of the