Analytical Solutions and Differential Equations

The dynamical problem types specify the dynamical models that are the nonlinear transformation of the NLME model. There are two major types of dynamical models: analytical models and DEProblems. An analytical model is a small differential equation with an analytical solution. This analytical solution is used by the solvers to greatly enhance the performance. On the other hand, DEProblem is a specification of a differential equation for numerical solution by DifferentialEquations. This is used for specifying dynamical equations which do not have an analytical solution, such as many nonlinear ordinary differential equations (ODEs), or the myriad of differential equation types supported by DifferentialEquations, such as delay differential equations (DDEs) and stochastic differential equations (SDEs).

Analytical Solutions

Analytical problems are a predefined ODE with an analytical solution. While limited in flexibility, the analytical solutions can be much faster for simulation and estimation. In the @model DSL, an analytical solution is declared by name. For example:

@dynamics Central1

declares the use of Central1 dynamics. Analytical solutions have preset names which are used in the internal model. These parameters must be given values in the @pre block.

`Central1`

Pumas.Central1 — Type

Central1

An analytical model for a one compartment model with dosing into Central. Equivalent to

Central' = -CL/Vc*Central

where clearance, CL, and volume, Vc, are required to be defined in the model through @param, @covariates, or @pre.

`Central1Periph1`

Pumas.Central1Periph1 — Type

Central1Periph1

An analytical model for a two-compartment model with a central compartment, Central and a peripheral compartment, Peripheral. Equivalent to

Central'    = -(CL+Q)/Vc*Central + Q/Vp*Peripheral
Peripheral' =       Q/Vc*Central - Q/Vp*Peripheral

where clearance, CL, and volumes, Vc and Vp, and distribution clearance, Q, are required to be defined in the model through @param, @covariates, or @pre.

`Central1Periph2`

Pumas.Central1Periph2 — Type

Central1Periph2

An analytical model for a two-compartment model with a central compartment, Central and a peripheral compartment, Peripheral. Equivalent to

Central'    = -(CL+Q1+Q2)/Vc*Central + Q1/Vp1*Peripheral1 + Q2/Vp2*Peripheral2
Peripheral1' =     Q1    /Vc*Central - Q1/Vp1*Peripheral1
Peripheral2' =        Q2 /Vc*Central                      - Q2/Vp2*Peripheral2

where clearance, CL, and volumes, Vc and Vp, and distribution clearance, Q, are required to be defined in the model through @param, @covariates, or @pre.

`Central1Meta1`

Pumas.Central1Meta1 — Type

Central1Meta1

An analytical model for a model with a central compartment, Central, and a metabolite, Metabolite. Equivalent to

Central'    = -(CL+CLfm)/Vc*Central
Metabolite' =       CLfm/Vc*Central  - CLm/Vm*Metabolite

where clearances (CL and CLm) and volumes (Vc and Vm) and formation clearance of metabolite CLfm are required to be defined in the model through @param, @covariates, or @pre.

`Central1Periph1Meta1`

Pumas.Central1Periph1Meta1 — Type

Central1Periph1Meta1

An analytical model for a two compartment model with a central compartment, Central, with a peripheral compartment, Peripheral, and a metabolite compartment, Metabolite. Equivalent to

Central'     = -(CL+Q+CLfm)/Vc*Central + Q/Vp*CPeripheral
CPeripheral' =            Q/Vc*Central - Q/Vp*CPeripheral
Metabolite'  = -CLm/Vm*Metabolite + CLfm/Vc*Central

where clearances (CL and CLm) and volumes (Vc, Vp and Vm), distribution clearance (Q), and formation clearance of metabolite CLfm are required to be defined in the model through @param, @covariates, or @pre.

`Central1Periph1Meta1Periph1`

Pumas.Central1Periph1Meta1Periph1 — Type

Central1Periph1Meta1Periph1

An analytical model for a two compartment model with a central compartment, Central, with a peripheral compartment, Peripheral, and a metabolite compartment, Metabolite, with a peripheral compartment, MPeripheral. Equivalent to

Central'     = -(CL+Q+CLfm)/Vc*Central + Q/Vp*CPeripheral
CPeripheral' =            Q/Vc*Central - Q/Vp*CPeripheral
Metabolite'  = -(CLm+Qm)/Vm*Metabolite + Qm/Vmp*MPeripheral + CLfm/Vc*Central
MPeripheral' =        Qm/Vm*Metabolite - Qm/Vmp*MPeripheral

where clearances (CL and CLm) and volumes (Vc, Vp, Vm and Vmp), distribution clearances (Q and Qm) and formation clearance of metabolite CLfm are required to be defined in the model through @param, @covariates, or @pre.

`Depots1Central1`

Pumas.Depots1Central1 — Type

Depots1Central1

An analytical model for a one compartment model with a central compartment, Central, and a depot, Depot. Equivalent to

Depot'   = -Ka*Depot
Central' =  Ka*Depot - CL/Vc*Central

where absoption rate, Ka, clearance, CL, and volume, Vc, are required to be defined in the model through @param, @covariates, or @pre.

`Depots1Central1Periph1`

Pumas.Depots1Central1Periph1 — Type

Depots1Central1Periph1

An analytical model for a two-compartment model with a central compartment, Central, a peripheral compartment, Peripheral, and a depot Depot. Equivalent to

Depot'      = -Ka*Depot
Central'    =  Ka*Depot - (CL+Q)/Vc*Central + Q/Vp*Peripheral
Peripheral' =                 Q/Vc*Central - Q/Vp*Peripheral

where absorption rate, Ka, clearance, CL, and volumes, Vc and Vp, and distribution clearance, Q, are required to be defined in the model through @param, @covariates, or @pre.

`Depots2Central1`

Pumas.Depots2Central1 — Type

Depots2Central1

An analytical model for a one compartment model with a central compartment, Central, and two depots, Depot1 and Depot2. Equivalent to

Depot1'  = -Ka1*Depot1
Depot2'  = -Ka2*Depot2
Central' =  Ka1*Depot1 + Ka2*Depot2 - CL/Vc*Central

where absorption rates, Ka1 and Ka2, clearance, CL, and volume, Vc, are required to be defined in the model through @param, @covariates, or @pre.

When using this model during simulation or estimation, it is preferred to have 2 dosing rows for each subject in the dataset, where the first dose goes into cmt =1 (or cmt = Depot1) and the second dose goes into cmt=2 (or cmt=Depot2). Central compartment gets cmt=3 or (cmt = Central). e.g.

ev = DosageRegimen([100,100],cmt=[1,2]) s1 = Subject(id=1, events=ev)

`Depots2Central1Periph1`

Pumas.Depots2Central1Periph1 — Type

Depots2Central1Periph1

An analytical model for a two-compartment model with a central compartment, Central, a peripheral compartment, Peripheral, and two depots Depot1 and Depot2. Equivalent to

Depot1'      = -Ka1*Depot1
Depot2'      = -Ka2*Depot2
Central'    =  Ka1*Depot1 + Ka2*Depot2 - (CL+Q)/Vc*Central + Q/Vp*Peripheral
Peripheral' =                 Q/Vc*Central - Q/Vp*Peripheral

where absorption rates, Ka1 and Ka2, clearance, CL, and volumes, Vc and Vp, and distribution clearance, Q, are required to be defined in the model through @param, @covariates, or @pre.

Differential Equations `DEProblem`

DEProblems are types from DifferentialEquations which are used to specify differential equations to be solved numerically via the solvers of the package. In the @model interface, the DEProblem is set to be an ODEProblem defining an ODE. The models are defined by writing each of the differential equations in the system, e.g.

@dynamics begin
    Central' = -(CL / Vc) * Central
end

for a simple one compartment model and

@dynamics begin
    Depot' = -Ka * Depot
    Central' = Ka * Depot - (CL / Vc) * Central
end

for a one compartment model with first order absorption.

Default solver

Dynamical models that do not have an analytical solution have to be integrated numerically. There are a myriad of solvers with different properties and performance under different conditions. In Pumas the default solver is

AutoVern7(
    Rodas5P(
        autodiff = ADTypes.AutoForwardDiff(; chunksize = 1),
        linsolve = LinearSolve.GenericLUFactorization(),
    ),
)

The AutoX(Y) signature where X and Y are names of integrators means that X will be used unless stiffness is detected and then Y will be used. In both simobs and fit (or any other function where it may be relevant to set diffeq_options) it is possible to change these defaults using the diffeq_options keyword argument with an alg element. For example, it is possible to fit using Vern7 only using the syntax

results = fit(model, population, param, FOCE(); diffeq_options = (; alg = Vern7()))

This should only be done if the user is certain that stiffness could not be detected even at extreme parameter values, but it would in principle remove a little bit of overhead if the model is indeed not stiff.