Search the PBPK Model Repository

Quickly find freely available drug and population models in our PBPK model repository.

The models provided have been collated from published examples which authors have shared in our Published Model Collection or developed as part of various global health projects in our Global Health Collection. This search facility searches both model collections simultaneously.

To contribute published user compound and/or population files, upload your files here: Upload Model Files

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

|<

<

1

2

3

4

5

6

7

8

9

10

11

12

13

14

>

>|

Found 126 Matches

TB Compound File Collection

V16 TB compound files and associated manuscript, as well as South African TB population file.

Ceftibuten_V14R1_AstraZeneca_20200327

Ceftibuten for pediatric predictions.

CrohnsDiseasePopulation_V19R1_UniversityOfManchester_20230322

This model is a Crohn's disease (CD) population model for adult patients, created in Simcyp version 19. Alteration of GIT parameters and other physiological parameters like albumin level incorporated based on literature available data of inactive state of CD. The demographic and formulation information are based on the published in vivo pharmacokinetics data of midazolam and budesonide in CD patients. The model details and results are published in 2022 DOI: 10.1007/s40262-022-01169-4 

Chlorpromazine

Brand Name(s) include: Thorazine, Largactil, Ormazine

Indication: Schizophrenia, manic-depression

Drug Class: Conventional anitpsychotic

Date Updated: March 2024

The model at-a-glance

  Absorption Model

  • First-Order, fa and ka predicted by Caco-2 data

  Volume of Distribution

  • Minimal (optimized to IV data)

  Route of Elimination

  • fmCYP3A4 = 80, fmCYP2D6 = 20

  Perpetrator DDI

  • Inhibition of CYP2D6

  Validation

  • Model performance was verified in healthy volunteers and psychiatric patients. Two clinical studies with IV administration (7 to 10 mg) and ten clinical studies with oral administration (25 to 200 mg) were used for model verification. Simulations for eight of eleven clinical studies with a reported AUC were within 2-fold of the observed value.
  • The fmCYP1A2 was verified through simulations of chlorpromazine in smokers and non-smokers. The fmCYP2D6­ was verified through simulations of chlorpromazine coadministered with and without quinidine.

  Limitations

  • Model was developed using 10 mg IV and 100 mg PO. Use of 7-10 mg IV dosing and oral doses of 100-200 mg were verified, but predictions of oral doses <100 mg PO are overestimated.
  • Model does not include P-gp efflux.
  • Model is not verified for use in perpetrator DDI simulations with CYP2D6 substrates.

 

|<

<

1

2

3

4

5

6

7

8

9

10

11

12

13

14

>

>|