Biology Reference
In-Depth Information
experimental data. The residuals appear random, and there are no
outliers.
1
Currently, an ''automatic'' multiparameter deconvolution technique is
available (titled AUTODECON) that is maximally assumption-free and
implements a rigorous statistical test for the existence of secretion
events. In addition, it eliminates the subjective nature of the earlier
algorithms by automatically inserting, and subsequently testing, the
significance of presumed secretion events. This automatic algorithm is a
combination of three modules; a parameter fitting module that performs
weighted nonlinear least-squares parameter estimation, an insertion
module that determines the location for and adds presumed secretion
events, and the triage module that removes secretion events that are
not statistically significant. We refer the reader to the references below
for a comprehensive description.
0.1
2
0
2
0.1
0.01
0.001
Before we move on, we reiterate that in this chapter we focused on
methods using hormone concentrations measured in the blood serum
to determine whether the secretion events are periodic or pulsatile
and to reconstruct the secretion levels from experimental data.
Sophisticated mathematical methods, such as convolution-
deconvolution techniques, are required in this case because of the
ongoing physiological elimination of the hormone. An alternative
approach would be to consider the factors impacting hormone secretion
and model the network of interactions in order to obtain the secretion
profile of the hormone in question. We consider such models in the
next chapter.
200
400
600 800 1000 1200 1400
Minutes
FIGURE 9-25.
Analysis of the growth hormone data in Figure 9-
6 by multiparameter deconvolution predicted 24
secretion events with HL ¼ 20.4 minutes, C(0) ¼
1.42, S 0 ¼ 0.00122, and Secretion SD ¼ 9.7
minutes. The concentrations and secretion rates
are shown on a logarithmic scale to enable
viewing of the diminutive secretion events.
REFERENCES
Box, G., Jenkins, G. M., & Reinsel, G. (1994). Time series analysis: Forecasting and
control (3rd ed.). Upper Saddle River, NJ: Prentice Hall.
Grafakos, L. (2004). Classical and modern Fourier analysis. Upper Saddle River, NJ:
Prentice Hall.
Guyton, A. C., & Hall, J. E. (2005). Textbook of medical physiology (11th ed.). New
York: Saunders/Elsevier.
Jansson, P. A. (1984). Deconvolution with applications in spectroscopy. New York:
Academic Press.
Johnson, M. L., & Veldhuis, J. D. (1995). Evolution of deconvolution analysis as a
hormone pulse detection method. Methods in Neuroscience, 28, 1-24.
Johnson, M. L., Virostko, A., Veldhuis, J. D., & Evans, W. S. (2004).
Deconvolution analysis as a hormone pulse-detection algorithm. In Johnson,
M. L., & Brand, L. (Eds.), Methods in Enzymology (vol. 384, pp. 40-53).
New York: Academic Press.
Urban, R. J., Evans, W. S., Rogol, A. D., Kaiser, D. L., Johnson, M. L., & Veldhuis,
J. D. (1988). Contemporary aspects of discrete peak detection algorithms: I. The
paradigm of the luteinizing hormone pulse signal in men. Endocrine Reviews, 9,
3-37.
Search WWH ::




Custom Search