5.7 Werhli AV. Comparative evaluation of reverse engineering gene regulatory networks with relevance networks, graphical Gaussian models and Bayesian networks.
See related topics and documents
5.7.1 Method comparison
5.7.2 (for microarrays)
5.7.3 (1) Relevance networks (RNs):
Pairwise association scores independent of the remaining network
5.7.4 (2) graphical Gaussian models (GGMs):
Undirected graphical models with constraint-based inference
5.7.5 (3) Bayesian networks (BNs):
Directed graphical models with score-based inference.
5.7.6 BN > GGM > RN
BN: higher performance cost than GGM
..but pays off for perturbation analysis