Bayesian probabilistic network modeling from multiple independent replicates

BMC Bioinformatics

Table 1 Study 1 posterior probabilities

	C⁽¹⁾	$R_{1}^{(1)}$	$R_{2}^{(1)}$	$R_{3}^{(1)}$
$p_{1}^{(1)} - p_{2}^{(1)} - p_{3}^{(1)}$	1.0	1.0	0.9744	1.0
$p_{4}^{(1)} - p_{5}^{(1)} - p_{6}^{(1)}$	1.0	1.0	1.0	1.0
$p_{7}^{(1)} - p_{8}^{(1)} - p_{9}^{(1)}$	1.0	1.0	1.0	1.0
$p_{10}^{(1)} - p_{11}^{(1)} - p_{12}^{(1)}$	1.0	1.0	0.993	1.0
TFP Avg	0.0376	0.0246	0.0641	0.0346
TLFP Avg	0.0008	0.0008	0.0047	0.0017

For the first study, the posterior connectivity probabilities for the four signal connected blocks are shown. For the composite model and the single rep models, all of these are quite high. The last two rows indicate the posterior probabilities (raw average and average via initial log odds transformation) over all false possible threesomes of vertices. These average posterior probabilities and average log odds are quite low.

ISSN: 1471-2105