In Exploration 2 it was found that for a short in vivo hyperthermic pulse can provide crude information about the development stage and processes it has perturbed.  Exploration 4 examines the possibility of connecting genotype to cellular or organismal phenotypes through the intervening chemotype.  The chemotype of a system consists of its components and their internal and interactional reaction rates.  The behavior of the system over time is determined by the initial state of the components and the subsequent reactions.

      However, even if we had measurements of what the transient changes in temperature were for this or any other method of externally inducing heat change, it would be difficult to apply this knowledge into any quantitative modeling.  However, various in vitro experimental setups have been found to mimic particular aspects of development, and it may be possible to get realistic estimates of various reaction rates.

      It might be thought that with 22,000 genes producing products that interact with each other in myriad ways, any such approach would be defeated by the complexity of the problem.  However, I think there are factors that may mitigate such pessimism.

      As described in Exploration 4, systems in living beings may be constrained to be sloppy.  A sloppy system is defined as one in which the precision to which each of its parameters must be constrained for the system to work properly (or at least adequately) is broadly distributed.  In particular, only a minority of the parameters are constrained to a narrow range.  Precise knowledge of just these parameters may be good enough to model the actual behavior.

      In Background 1 evidence is discussed that the neck of mammals has a more modular organization than that of birds.  When comparing two systems that accomplish the same or similar goals, the more modular will show fewer interactions between the components.  In terms of coping with various degrees of heterothermy, in vivo development in model systems of other vertebrates, such as chicken, Xenopus, or zebrafish is much more accessible to experimental manipulation and measurements.  Comparing data from these to mammalian in vitro data may reveal evidence of the evolution of such modularity.

     In Exploration 4 the protein, KaiC, that plays the central role in the circadian clock mechanism in cyanobacteria is discussed. Wild type KaiC goes through a complicated cycle involving auto kinase (phosphorylation) and phosphatase activity at two sites, ATPase activity at a different site, and confirmational changes, with a period of period near 24 hours independently of temperature from 25 to 35 degrees C.  The rate of enzymatic reactions typically increases 2-3-fold over this range.  In KaiC mutants, the cycle time varies inversely with the ATPase rate.  For most of these mutants the rate is no longer independent of temperature, either increasing or decreasing.  Mutations that change the amino acid at a particular residue also change cycling time but retain the temperature independence.

      It has been possible to isolate the many of the intermediate states of wild type and mutant proteins and determine their 3D structure by xray crystallography, leading to increasing understanding of the transition rates between the states.  Studies of wild type KaiC with elastic neutron scattering indicate that parts of the molecule are unusually stiff, showing less thermal vibration than expected, which may be related to the temperature independence of cycling.

      It has been said that one of the most impressive results using AI techniques has been the ability to reliably predict protein 3D structure from the amino acid sequence.  Using a combination of in vivo, in vitro, and in silico methods, it may be possible to rapidly increase knowledge of enzymatic reactions and their relationship to the genotype.

      The following example shows the value of such systems in understanding differences in the segmentation clock among mammalian species using a stem cell "zoo."

​

​