I learned of neural nets from a friend who is a commodities broker and
uses them to
predict trading patterns according to the weights that he assigns to each
indicator. I
was wondering if similar predictive abilities might be afforded by neural
nets when one
applies them to create a model for hormone actions in the body. I've
been running this
theory around in my head and am wondering if this is at all feasible.
I first started thinking about it when I learned of the comparison of
hormone signal
pathways to neural pathways and how each has an effect on the target
neuron/organ
depending on the concentration/strength of the signal, such as actual
molar concentration,
ability of the hormone to initiate a 2nd messenger response within a
cell, or strength
of signal due solely to the action of the hormone itself. This compares
nicely with
neural signals, as far as I can tell. I'm just an undergrad, after all
:) I would think
that the model would be extremely simplistic, but it would be a start of
something that
I feel would be of incredible help, since endocrinology is still such an
infant field.
I would be particularly grateful if someone could direct me to a science
periodical
that might discuss this possibility, a researcher who is exploring this
idea, or just
give me feed-back as to the feasibility of this idea.
I am about to start making application to grad schools, and this is a
topic that I
feel I would like to explore, so if anyone has any additional information
on schools
that would be formidable in this area, please let me know.
Christopher Schultz
2600 Rio Grande
Austin, Texas 78705
cschultz at utxvms.cc.utexas.edu
(home) 512.477.5475
(lab) 512.471.3127
