Table 2. Results from the optPBN toolbox for case study 2 compared to the original network.

Optimisation results
Exp|rules	PI	PT	PI&∼PT	Ø
A,D	0.4602	0.1344	0.3607	0.0447
All (A to D)	0.6041	0	0.3959	0

The table shows the results of optimisation for two datasets: 1) containing measurement data from experiments A and D 2) containing all measurement data from experiments A, B, C and D. Four different Boolean rules are applied as follows: only activation from PI3K (PI), only activation from PTEN (PT), activation from PI3K and inhibition from PTEN (PI&∼PT), and no interaction from either PI3K or PTEN and output is fixed to 0 (Ø). The selection probabilities inferred from all experiments agree well with the selection probabilities of the original network. The dataset consisting of the measurement from only experiments A and D is insufficient to reconstruct the original network.