The genetic expression curve could be measured by varying a selected
set of environmental parameters, and the effect of combinations of these
parameters could define the growth curve you're describing.
In a laboratory, very careful control of all the variables (such as
total moisture, total light and spectrum exposure, specific chemicals)
with a computer controlled system could mathematically characterize
the effect you're describing. Every instance of the genetically identical
individual could be measured with sensors that report on these variables,
or high accuracy mesurement dispensers for water, chemicals, etc,
could very precisely control the variables to get the variation you
describe versus the controlled variables.
Next, consider two families of plants, with the members of each family
genetically identical to each other, but different from the other family
in genetically known ways. The same experiment should show a
different expression function for the two families. The differences
in growth curve could be attributed to the genetic differences solely.
Then, consider two families that are genetically identical except for
one single gene - family A has one version of the gene, and family
B has a different version. Conducting the same experiment on the
two families can result in measured growth curves for the polymorphism
of the single gene.
So, with N genetic variants, there are 2^N differential experiments
that could be run for identifying every combination of genetic variants
and how they interact. Of course, 2^N is a big number when you
consider that yeast cells have about 6,000 genes (not just variants),
e. coli have about 4,500, and people have about 60,000. Those
kinds of numbers are clearly astronomical, and not feasible to
So the only practical thing to do with this thought experiment is to
identify very specific genes that might be important, and run a few
limited experiments with minor variations.
The immediate question is: what genes are important enough to
participate in these experiments?
Yeast and e. coli might have valuable things to teach us, and they
could certainly be raised efficiently in an automated way compared
to fruitflies, potatoes, or mice.
Are there known genes that would be very useful to measure in this
way among the yeast, e. coli, or other "easy" organisms?