Biology Reference
In-Depth Information
(-)
(+)
GHRH
SRIF
Elimination
Elimination
(-)
(+)
GH
D
Elimination
FIGURE 10-5.
Consensus GH network.
awake, or nutrient-replete humans. Figure 10-6 schematizes a typical GH
secretion profile in the adult male rat (for real in vivo data see, for
example, Lanzi and Tannenbaum [1992]). A model that describes the
dynamics of GH concentration should reproduce the specifics of this
pattern and suggest possible explanations for the mechanisms that are
responsible for the observed dynamic behavior.
400
300
200
In this chapter, we discuss a general modeling approach to studying
various mechanisms of hormone release control, similar to the GH
network described above. The following three phases are fundamental to
the modeling effort:
100
0
2
4
6
8
Time [h]
1. Data analysis and exploration of the specifics of hormone concentra-
tion time series, such as pulse detection, analysis of the variability
and orderliness, determining the baseline secretion and half-life,
and detecting the frequency of oscillations. As a result of the data
analyses, selected experimental outcomes and hormone profile
specifics are targeted for explanation in the modeling effort.
Many aspects of this phase were described in Chapter 9.
FIGURE 10-6.
Typical schematized adult male rat GH secretion
profiles.
2. Formal network design presenting an intuitive outline of the
system's functional connectivity. This phase is based on
analysis of available data from phase 1 and on the interactions
between key system elements. The information is organized as a set
of nodes and conduits in a hypothetical formal endocrine network.
3. Dynamic modeling of the formal network. The formal endocrine
network from phase 2 is interpreted as a dynamic system
and described with a set of coupled ordinary differential
equations (ODEs). They give the time derivative of each network
node and approximate all system positive and negative dose-
responsive control links. The parameters in the ODEs must have
a clear physiological meaning and are determined by comparing
the model output with the phase 1 data.
The outcome of the modeling effort addresses the question of whether
the design of the formal endocrine network is a good approximation of
