Information Technology Reference
In-Depth Information
system emerges that is sufficient to implement a recognition system [3, 4, 2]. The
properties of such genetic regulatory networks are investigated.
7.2 Model and Methods
This section describes the structure, function, and equations of the genetic regula-
tion model.
7.2.1 Basic Assumptions
The role of a gene promoter is to measure the amount of product available and deter-
mine gene expression levels. The most important assumption, on which this model
builds upon, is that each promoter aims to produce a fixed amount of product. If too
much product is consumed, the promoter signals more product must be expressed. If
too little is consumed, the promoters signal less to be expressed. A gene that affects
multiple products is regulated by those products. Thus, every input-output relation
is regulated by feedback.
A classifier based on feedback can be surprisingly powerful [3, 4]. This structure
maintains its simplicity in large networks but can still make complex recognition
decisions based on distributed processing.
7.2.2 Model Structure
The proposed tight association between genes and products and promoters is de-
picted in Fig. 7.1.
Production Feedback
Genes: y
i
y
1
y
3
y
2
y
4
Promoters:
f
i
f
2
f
1
Products:
x
i
x
1
x
2
Fig. 7.1: Self-regulation. If
y
1
and
y
2
affect
x
1
then
f
1
monitors
x
1
and regulates
y
1
and
y
2
. Similarly if
y
1
,
y
2
,
y
3
, and
y
4
affect
x
2
then
f
2
monitors
x
2
and regulates
y
1
,
y
2
,
y
3
, and
y
4
.
