Processes 2018, 6,56
and corresponding simulation to explore this response. The model is based on two other
Analogous-mechanism Models, (1) a model of lipoprotein metabolism and kinetics and
(2) a model of RG7232 pharmacokinetics. They are combined into a single simulation.
The linked simulation goes further by additionally representing the hypothesis that the
affinity of low-density lipoprotein (LDL) particles to LDL receptors are dependent on
particle size or density. This hypothesis is implemented as a modified elimination rate.
The resulting model describes temporal concentrations in two-compartments as coupled
ordinary differential equations that are solved using the SimBiology toolbox of MathWorks.
The simulation model is “analogous” in the sense that the proposed density-dependent
elimination rate and compartmentalization is an analogy to chemical kinetics and chemical
engineering. Parameters are estimated using a Bayesian approach that updates the parameter
values from model components using the Matlab Global Optimization toolbox of MathWorks.
The implementations simulate output from the linked Analogous-mechanism Model as if it
were real.
Example V.2: More than 40% of astronauts who participate in long-duration missions return
with ophthalmic changes similar to idiopathic intracranial hypertension. Experts posited
that a microgravity-induced cephalic fluid shift elevates intracranial pressure (ICP).
Feola et al. [32] hypothesized that elevated ICP would alter the peak strain environment in the
optic nerve head (ONH) to cause tissue remodeling that may be contributing to the observed
ophthalmic changes. They also suspected that variations in intraocular pressure (IOP) and
mean arterial pressure (MAP) would affect the biomechanical strain in the OHN tissues.
To explore that explanation, they implemented a finite element Analogous-mechanism
Model in which a simulated structural mechanism is strongly analogous to (functions as
an analog of) the ocular structure. The geometry of the analog was based on established
ocular biomechanics research and it included representing coarse grain features of tissue
structures known to play a significant role in the observed ophthalmic changes: sclera,
preliminary neural tissue, lamina cribrosa, central retinal vessel, dura mater and pia mater of
the optic nerve sheath. Furthermore, an annular ring was incorporated around the scleral
canal to account for the circumferential alignment of the scleral collagen fibers around
the ONH. The open source package Gmsh (V2.8.3) was used to generate the 3D finite
element geometry and mesh and open source FE solver FEBio (V2.0) was used to solve
for all simulations. The authors used Latin hypercube sampling of biologically plausible
regions of parameter space to simulate biomechanical responses of their analog eye structure
to various combinations of simulated ICPs, as well as varying IOP, MAP, and simulated
tissue mechanical property conditions. Execution results showed that chronically elevated
ICP coupled with interindividual differences in simulated optic nerve head mechanical
properties could influence the risk for experiencing extreme optic nerve strains. The authors
inferred that individuals with both soft optic nerve or pia mater and elevated ICP would be
especially at risk.
Example V.3: Rosiglitazone is a PPARγ agonist, one of several approved insulin sensitizers
used to treat diabetes. Despite being on the market for over a decade, the drug continues
to be studied in the lab to understand the mechanism of action of this class of molecule.
In Goto-Kakizaki rats, which are a rodent model of early-developing, non-obese type-2
diabetes, Gao and Jusko [30] show that rosiglitazone decreases glucose levels. To simulate
how the insulin/glucose regulation might work, they built a feedback model—glucose
stimulating insulin production and insulin increasing glucose consumption. The model
is analogous to other simple feedback systems, without specifying the actual, detailed,
biological mechanism (e.g., intermediate steps) for glucose/insulin co-regulation. The model
also incorporates two pharmacodynamic effects of rosiglitazone that impact this feedback
193
and corresponding simulation to explore this response. The model is based on two other
Analogous-mechanism Models, (1) a model of lipoprotein metabolism and kinetics and
(2) a model of RG7232 pharmacokinetics. They are combined into a single simulation.
The linked simulation goes further by additionally representing the hypothesis that the
affinity of low-density lipoprotein (LDL) particles to LDL receptors are dependent on
particle size or density. This hypothesis is implemented as a modified elimination rate.
The resulting model describes temporal concentrations in two-compartments as coupled
ordinary differential equations that are solved using the SimBiology toolbox of MathWorks.
The simulation model is “analogous” in the sense that the proposed density-dependent
elimination rate and compartmentalization is an analogy to chemical kinetics and chemical
engineering. Parameters are estimated using a Bayesian approach that updates the parameter
values from model components using the Matlab Global Optimization toolbox of MathWorks.
The implementations simulate output from the linked Analogous-mechanism Model as if it
were real.
Example V.2: More than 40% of astronauts who participate in long-duration missions return
with ophthalmic changes similar to idiopathic intracranial hypertension. Experts posited
that a microgravity-induced cephalic fluid shift elevates intracranial pressure (ICP).
Feola et al. [32] hypothesized that elevated ICP would alter the peak strain environment in the
optic nerve head (ONH) to cause tissue remodeling that may be contributing to the observed
ophthalmic changes. They also suspected that variations in intraocular pressure (IOP) and
mean arterial pressure (MAP) would affect the biomechanical strain in the OHN tissues.
To explore that explanation, they implemented a finite element Analogous-mechanism
Model in which a simulated structural mechanism is strongly analogous to (functions as
an analog of) the ocular structure. The geometry of the analog was based on established
ocular biomechanics research and it included representing coarse grain features of tissue
structures known to play a significant role in the observed ophthalmic changes: sclera,
preliminary neural tissue, lamina cribrosa, central retinal vessel, dura mater and pia mater of
the optic nerve sheath. Furthermore, an annular ring was incorporated around the scleral
canal to account for the circumferential alignment of the scleral collagen fibers around
the ONH. The open source package Gmsh (V2.8.3) was used to generate the 3D finite
element geometry and mesh and open source FE solver FEBio (V2.0) was used to solve
for all simulations. The authors used Latin hypercube sampling of biologically plausible
regions of parameter space to simulate biomechanical responses of their analog eye structure
to various combinations of simulated ICPs, as well as varying IOP, MAP, and simulated
tissue mechanical property conditions. Execution results showed that chronically elevated
ICP coupled with interindividual differences in simulated optic nerve head mechanical
properties could influence the risk for experiencing extreme optic nerve strains. The authors
inferred that individuals with both soft optic nerve or pia mater and elevated ICP would be
especially at risk.
Example V.3: Rosiglitazone is a PPARγ agonist, one of several approved insulin sensitizers
used to treat diabetes. Despite being on the market for over a decade, the drug continues
to be studied in the lab to understand the mechanism of action of this class of molecule.
In Goto-Kakizaki rats, which are a rodent model of early-developing, non-obese type-2
diabetes, Gao and Jusko [30] show that rosiglitazone decreases glucose levels. To simulate
how the insulin/glucose regulation might work, they built a feedback model—glucose
stimulating insulin production and insulin increasing glucose consumption. The model
is analogous to other simple feedback systems, without specifying the actual, detailed,
biological mechanism (e.g., intermediate steps) for glucose/insulin co-regulation. The model
also incorporates two pharmacodynamic effects of rosiglitazone that impact this feedback
193