Systems Biology (Methods in Molecular Biology)

different sub-networks by acting on the sign of links or even form-

ing new links aiming to reach a more stable global condition.

Figure7 (top panel) shows a simple 9-node network imple-

mented in a Multi Agents System accounting for the arrangement

of both positive and negative correlations as well as for the homeo-

static regulation between regions (sub-networks A, B, C) through a

negative feedback mechanism. Thanks to the graphical facilities of

Netlogo™[9, 10] environment, the (de)activation signals traveling

from input node to output nodes are easily reproduced and visua-

lized in terms of time-dependent changes in the nodes’ size.

The mechanism in Fig.7, coupled to the underlying MAS

programming environment, reproduced some relatively sophisti-

cated homeostatic phenomena (seebelow), and its possible exten-

sion to the study of stability alterations of pathological significance

is foreseeable.

(I)

A

B C

(II)

A

B C

t (a.u.)

Node Energy

-0.5 3.5

0 0.5 1

0 1

6

1

0

6

Node 0

Node 1

Node 6

Fig. 7Top: Dynamic features of a simple 9-node network. The direction of the link between nodes 0, 1 is
0 !1in(II) and 1!0in(I). The latter condition is the only one compatible with a global stability.Bottom:
The (I)!(II) and (II)!(I) transitions start att¼0 andt¼0.5, respectively. The graph describes the
consequential time-dependent changes in activity (size) of nodes 0, 1, and 6 (representative of the three
sub-networks) (modified from [32])

Multi-agent Simulations of Population Behavior... 317