Information Technology Reference
In-Depth Information
a)
B
A
C
A* = D
B* = A OR C
C* = A
D* = B AND C
b)
D
c)
D
A*
A
C*
0
0
0
0
1
1
1
1
B C D*
0 0 0
1 0 0
0 1 0
1 1 1
Fig. 4.5. A simple Boolean network. a) Wiring diagram. The edges represent regulatory
relationships, such that the state of the node at the endpoint of the edge depends on the state of the
node at the beginning of the edge. b) Logical Boolean rules. The * marks the future state of the
marked node. c) State look-up table for each node. The left column(s) represent the input state of
the regulating node(s), and the right column represents the output state of the node which is
connected by the incoming arrow to the regulating node(s).
Boolean functions to each node. The random selection of Boolean functions
can be performed once at the beginning or at each time step (Derrida et al. 1986)
by a machine learning algorithm. The dynamics of state transitions is not
dependent on the methods of selection of the functions mentioned above.
Algorithms such as REVEAL (REVerse Engineering Algorithm) (Liang et al.
1998) offer promising first steps towards large-scale network inference. In the
synchronous Boolean algorithm of REVEAL, dependencies between the
expressions of genes are calculated by using a quantitative information measure
so that the particular rule for a gene can describe its state transitions in the time-
series data.
A
C
B*
0
0
0
1
0
1
0
1
1
1
1
1
Search WWH ::




Custom Search