Skip to content

Models Reference

NeoPKPD provides a comprehensive library of validated pharmacokinetic (PK) and pharmacodynamic (PD) models for drug concentration and effect simulation.

Model Categories

  • Pharmacokinetic Models

    Compartmental models describing drug distribution and elimination

    PK Models

  • Pharmacodynamic Models

    Effect models linking concentration to response

    PD Models

Pharmacokinetic Models

Compartmental PK Models

Model Page Parameters Route Use Case
One-Comp IV Bolus → CL, V IV Simple IV kinetics
One-Comp Oral → Ka, CL, V Oral Simple oral drugs
Two-Comp IV → CL, V1, Q, V2 IV Distribution kinetics
Two-Comp Oral → Ka, CL, V1, Q, V2 Oral Oral with distribution
Three-Comp IV → CL, V1, Q2, V2, Q3, V3 IV Deep tissue
Transit Absorption → N, Ktr, Ka, CL, V Oral Delayed absorption
TMDD → Multiple IV Target-mediated disposition

Model Selection Guide

graph TD
    A[Start] --> B{Route?}
    B -->|IV| C{Distribution?}
    B -->|Oral| D{Absorption?}

    C -->|Mono-exponential| E[One-Comp IV]
    C -->|Bi-exponential| F[Two-Comp IV]
    C -->|Tri-exponential| G[Three-Comp IV]
    C -->|Nonlinear| H[Michaelis-Menten]

    D -->|First-order| I{Distribution?}
    D -->|Delayed| J[Transit Absorption]

    I -->|Simple| K[One-Comp Oral]
    I -->|Complex| L[Two-Comp Oral]

Pharmacodynamic Models

Effect Models

Model Page Parameters Type Mechanism
Direct Emax → E0, Emax, EC50 Direct Hyperbolic response
Sigmoid Emax → E0, Emax, EC50, γ Direct Hill equation
Effect Compartment → ke0, E0, Emax, EC50 Indirect Biophase equilibration
Disease Progression → Multiple Complex Disease dynamics

PD Model Selection Guide

graph TD
    A[Start] --> B{Temporal relationship?}
    B -->|Immediate| C[Direct Models]
    B -->|Delayed| D{Mechanism?}

    C --> E{Response shape?}
    E -->|Hyperbolic| F[Direct Emax]
    E -->|Steep/Sigmoid| G[Sigmoid Emax]

    D -->|PK-PD lag| H[Effect Compartment]
    D -->|Turnover| I{Drug effect?}

    I -->|Inhibition| J{Target?}
    I -->|Stimulation| K{Target?}

    J -->|Production| L[IDR Type I]
    J -->|Elimination| M[IDR Type II]

    K -->|Production| N[IDR Type III]
    K -->|Elimination| O[IDR Type IV]

Common Features

Dose Events

All models support flexible dosing:

# Single bolus
doses = [DoseEvent(0.0, 100.0)]

# Multiple doses
doses = [
    DoseEvent(0.0, 100.0),
    DoseEvent(12.0, 100.0),
    DoseEvent(24.0, 100.0)
]

# IV Infusion (1-hour)
doses = [DoseEvent(0.0, 100.0, 1.0)]

Observation Types

Observation Symbol Description
Concentration :conc Drug concentration in central compartment
Effect :effect Pharmacodynamic effect
Response :response Indirect response biomarker
Ce :ce Effect site concentration

Solver Options

# Standard solver for non-stiff problems
solver = SolverSpec(:Tsit5, 1e-10, 1e-12, 10_000_000)

# For stiff problems (e.g., rapid binding)
solver = SolverSpec(:Rosenbrock23, 1e-8, 1e-10, 1_000_000)

# High accuracy for validation
solver = SolverSpec(:Vern9, 1e-14, 1e-14, 100_000_000)

Model Implementation

Type Hierarchy

# Abstract types
abstract type ModelKind end
abstract type AbstractParams end

# PK model kinds
struct OneCompIVBolus <: ModelKind end
struct TwoCompIVBolus <: ModelKind end
# ...

# Parameter types
struct OneCompIVBolusParams <: AbstractParams
    CL::Float64
    V::Float64
end

Creating Custom Models

See the source code in packages/core/src/models/ for examples of extending NeoPKPD with custom models.

Next Steps