Bayesian probabilistic network modeling from multiple independent replicates

BMC Bioinformatics

Table 8 Study 5 posterior probabilities

	C⁽⁵⁾	$R_{1}^{(5)}$	$R_{2}^{(5)}$	$R_{3}^{(5)}$
$p_{1}^{(5)} - p_{2}^{(5)} - p_{3}^{(5)}$	1.0	1.0	0.0016	1.0
$p_{4}^{(5)} - p_{5}^{(5)} - p_{6}^{(5)}$	1.0	1.0	0.9874	0.9819
$p_{7}^{(4)} - p_{8}^{(4)} - p_{9}^{(4)}$	1.0	1.0	1.0	1.0
$p_{10}^{(5)} - p_{11}^{(5)} - p_{12}^{(5)}$	1.0	1.0	1.0	0.3477
TFP Avg	0.0643	0.0418	0.0854	0.0613
TLFP Avg	0.0019	0.0023	0.0017	0.0017

For the fifth study, the posterior connectivity probabilities for the four signal connected blocks are shown, where $R_{1}^{(5)}, R_{2}^{(5)}, R_{3}^{(5)}$ are generated with three different ρ values, 0.9, 0.82 and 0.7, respectively. For the composite model and the single rep models, all of these are very high. The last two rows indicate the very low average posterior probability over all non-signal threesomes of vertices.

ISSN: 1471-2105