Processes 2018, 6,56
system: enhancing insulin sensitivity (i.e., increasing the rate of insulin-dependent loss
of glucose) and inhibiting glucose production. As with many models of pharmacology,
the pharmacokinetic part uses an idealized one-compartment model to fit observed drug
absorption and loss. The simulation is implemented using coupled ODEs, plus analytical
expressions for some of the molecules. Given its importance to diabetes and the system
under study, the component representing the time-dependent body weight of the rats was a
key variable being simulated along with the molecular components. Guided by experimental
measurements (of drug, glucose and insulin levels over time), the model was parameterized
for control, low dose and high dose rosiglitazone cases. The match between simulation output
and experiment measurements showed that the Analogous-mechanism Model explained
the observations sufficiently well. Using that model, the authors identified drug regimen
design principles: specifically, to enhance insulin sensitivity in the long term (>6 weeks),
a high-dose drug is needed continuously; neither lower-dose nor shorter-term treatment
succeeded in elevating the sensitivity.
Example V.4: Attempts to design and build synthetic cellular memory systems using
recombinases have thus far been hindered by a lack of validated computational models of a
plausible mechanism representing DNA recombination. The predictive capabilities of such
models are needed to reduce the number of iterative cycles required to align experimental
results with design performance requirements. Bowyer et al. [31] developed and validated
the first Simulation of an Analogous-mechanism Model for how DNA recombination
might occur. The models were constructed by extracting verified biological details from
an extensive review of the experimental literature and made use of a model analogy
with well-established reactions networks common to chemistry and chemical engineering.
Three essential biological details for which a consensus was lacking were included/excluded
from the simulations. The computational model consisted of a system of ODEs, each
representing the concentration of a distinct biological entity and model parameters that
were optimized via the use of genetic algorithms to refine parameter values but no
details on how the model was implemented were provided. Model predictions were
compared to experimental data to determine which set of details might represent the
most plausible mechanisms and thus serve as analogs of actual structural details by
which DNA recombination works. They found that including unidirectional (versus
bidirectional) excision, limiting recombinase directionality factor to monomeric form in
solution (versus dimer or tetramer) and integrase monomer (versus dimer) binding to
DNA produced the best model match to the data. Referring to Table 1, the contextual
location this Analogous-mechanism Model is implied but is not part of the implemented
computational model.
5.3. Group C: Using Computation to Support and Enhance Model Mechanisms
5.3.1. VI—Simulation of a Model Mechanism
The computational mechanisms used during simulation of an Analogous-mechanism Model
have nothing in common with referent mechanism’s spatiotemporal entities and activities within the
biological context. When a description of a Model Mechanism is available (III), it is feasible to change
that reality by striving to simulate an operating, concretized software (virtual) version of the Model
Mechanism. The research goal becomes twofold. (1) Create a discretized specification of the operating
Model Mechanism to guide development and instantiation of a virtual mechanism. Doing so requires
meeting this requirement: key portions of the virtual Model Mechanism operate during execution
as described in III and contribute to the simulation of Model Mechanism features. (2) Output and
measurements taken during simulations are qualitatively and quantitatively similar to measurements
of the target phenomenon.
194
system: enhancing insulin sensitivity (i.e., increasing the rate of insulin-dependent loss
of glucose) and inhibiting glucose production. As with many models of pharmacology,
the pharmacokinetic part uses an idealized one-compartment model to fit observed drug
absorption and loss. The simulation is implemented using coupled ODEs, plus analytical
expressions for some of the molecules. Given its importance to diabetes and the system
under study, the component representing the time-dependent body weight of the rats was a
key variable being simulated along with the molecular components. Guided by experimental
measurements (of drug, glucose and insulin levels over time), the model was parameterized
for control, low dose and high dose rosiglitazone cases. The match between simulation output
and experiment measurements showed that the Analogous-mechanism Model explained
the observations sufficiently well. Using that model, the authors identified drug regimen
design principles: specifically, to enhance insulin sensitivity in the long term (>6 weeks),
a high-dose drug is needed continuously; neither lower-dose nor shorter-term treatment
succeeded in elevating the sensitivity.
Example V.4: Attempts to design and build synthetic cellular memory systems using
recombinases have thus far been hindered by a lack of validated computational models of a
plausible mechanism representing DNA recombination. The predictive capabilities of such
models are needed to reduce the number of iterative cycles required to align experimental
results with design performance requirements. Bowyer et al. [31] developed and validated
the first Simulation of an Analogous-mechanism Model for how DNA recombination
might occur. The models were constructed by extracting verified biological details from
an extensive review of the experimental literature and made use of a model analogy
with well-established reactions networks common to chemistry and chemical engineering.
Three essential biological details for which a consensus was lacking were included/excluded
from the simulations. The computational model consisted of a system of ODEs, each
representing the concentration of a distinct biological entity and model parameters that
were optimized via the use of genetic algorithms to refine parameter values but no
details on how the model was implemented were provided. Model predictions were
compared to experimental data to determine which set of details might represent the
most plausible mechanisms and thus serve as analogs of actual structural details by
which DNA recombination works. They found that including unidirectional (versus
bidirectional) excision, limiting recombinase directionality factor to monomeric form in
solution (versus dimer or tetramer) and integrase monomer (versus dimer) binding to
DNA produced the best model match to the data. Referring to Table 1, the contextual
location this Analogous-mechanism Model is implied but is not part of the implemented
computational model.
5.3. Group C: Using Computation to Support and Enhance Model Mechanisms
5.3.1. VI—Simulation of a Model Mechanism
The computational mechanisms used during simulation of an Analogous-mechanism Model
have nothing in common with referent mechanism’s spatiotemporal entities and activities within the
biological context. When a description of a Model Mechanism is available (III), it is feasible to change
that reality by striving to simulate an operating, concretized software (virtual) version of the Model
Mechanism. The research goal becomes twofold. (1) Create a discretized specification of the operating
Model Mechanism to guide development and instantiation of a virtual mechanism. Doing so requires
meeting this requirement: key portions of the virtual Model Mechanism operate during execution
as described in III and contribute to the simulation of Model Mechanism features. (2) Output and
measurements taken during simulations are qualitatively and quantitatively similar to measurements
of the target phenomenon.
194