Qualification‌

Qualification framework

Introduction

The qualification framework enables an automated validation of various scenarios (use-cases) supported by the OSP platform. This technical framework is used, for example, to release, in full confidence, a new version of the OSP Suite by verifying automatically that an ever-growing list of scenarios is performing as expected.

A qualification scenario can be performed after an evaluation of the involved PBPK models has been done. A PBPK model evaluation only contains the healthy adult model development, and is divided into the following steps:

• PBPK model development and verification with observed data

• Model evaluation plan generation ("evaluation plan" = "qualification plan for one model")

• Evaluation report generation ("evaluation report" = "qualification report for one model")

The workflow of a PBPK model evaluation is similar to that of a PBPK model scenario qualification. A qualification scenario can be based on a single PBPK model or several models and is divided into the following steps:

• Scenario qualification (pure predictions) with observed data (e.g. DDI scenario, Enzyme ontogeny scenario)

• Qualification plan generation

• Qualification report generation

As a PBPK model evaluation workflow is similar to that of a qualification scenario, with the difference being model development and model application, respectively, the focus here will be on the scenario qualification workflow. In a first step, the qualification scenario is saved to a dedicated qualification repository on GitHub. This repository contains a detailed qualification plan that links and combines respective models and data describing the use case to qualify. The qualification plan consists of:

• PK-Sim project files (more precisely: PK-Sim project file snapshots)

• Descriptions of potential cross-dependencies between PK-Sim project files (if adequate) (e.g. it is possible to inherit building blocks or simulation parameters)

• Observed data sets (needed for model development and verification)

• Qualification scenario description text modules

• Detailed report settings related to the generation of charts and qualification measures.

Any file used in the qualification plan (e.g. PK-Sim projects, observed data sets, text modules etc...) can also be saved in an external repository and then conveniently referenced by the qualification plan.

In the next step, the Qualification Runner (stand-alone tool) processes the qualification plan, i.e. all project parts are exported and prepared for the Reporting Engine. The Reporting Engine provides a validated environment (currently implemented in MATLAB®, a transfer to R is in development) for model execution and generates tables and figures for the final qualification report. This report contains the evaluation of the individual PBPK models with observed data (i.e. standard goodness of fit plot, residuals-vs-time plot, visual predictive checks) and a comprehensive qualification of the specific use case assessing the predictive performance of the OSP suite by means of a predefined set of qualification measures and charts. The automated execution of the described workflow can be triggered to assess re-qualification, for example, when new data is available, after changes in model structure or parameterization, or when releasing a new version of the OSP Suite.

Creating a (re-)qualification plan part I

Creating a qualification report is similar to writing a scientific article: A report is written and structured in chapters, for example beginning with a short description of the scientific background of the scenario (use-case), followed by a brief methodological description (e.g. modeling strategy, available data used during model building (for model evaluation report), and underlying main assumptions) and the presentation of the qualification workflow results in the third section of the report.

The qualification plan orchestrates this process and defines how all the static and dynamic content will be combined into the final report document.

• “Static content”: Will be taken AS IS and inserted into the report without any further modifications.

• “Dynamic content”: Software must actively do something to produce expected results (e.g. create plots). This content may change between OSP versions in case of differences between the previous and new model structures/parameterizations.

Technically, a qualification plan is nothing more than a text file in JSON format (file extension: .json). You can use any plain text editor for creating and modification of such a file. However, it is much faster and easier to use a json editor (e.g. the free Visual Studio Code (VSCode); s. the section Creating a (re-)qualification plan part II: Tools for details). Note that many scripting environments (Matlab, R, etc...) also allow for the comfortable editing of JSON files.

Components of a (re-)qualification plan

Projects

Describes all projects used in a qualification scenario. Currently, only PK-Sim projects are supported. MoBi projects will be supported in the mid-term future.

• "Id": Whenever a project is referenced within a qualification plan: it happens via its project id. Any non-empty string can be defined by the author of a qualification plan as a project id (the only restriction: a project id must be unique within one qualification plan).

• "Path": path to a project snapshot. Can be defined:

  • Either in form of an URL of remote file (e.g. "https://github.com/Open-Systems-Pharmacology/Sufentanil-Model/releases/download/v1.1/Sufentanil.json"

  • or in form of a path to a LOCAL file (given relative to the location of current qualification plan), e.g. "Input/Itraconazole-Model.json")

• "BuildingBlocks": OPTIONAL, may be empty. List of inherited building blocks.

  The idea behind is: The use-case requires some building blocks (e.g. compound, individual, ...) to be exactly the same in one or more projects. Instead of modifying those projects by hand, the qualification plan can automate this action and ensure that building blocks are used consistently.

  • "Type": type of a building block (one of: "Compound", "Event", "Formulation", "Individual", "ObserverSet", "Population", "Protocol")

  • "Name": name of a building block (must be the same in both parent and child project)

  • "Project": Id of the parent project

  Example

  • Individual building block "European (P-gp modified, CYP3A4 36 h)" in the project "Itraconazole-Midazolam-DDI" will be overwritten by the Individual building block with the same name from the project "Midazolam" (if there is no individual with the same name in the "Midazolam" project: execution of the qualification plan will stop with an error)

  • Compound building block "Midazolam" in the project "Itraconazole-Midazolam-DDI" will be overwritten by the Compound building block with the same name from the project "Midazolam" (if there is no Compound with the same name in the "Midazolam" project: execution of the qualification plan will stop with an error)

• "SimulationParameters": OPTIONAL: List of inherited simulation parameters (i.e. parameters that are not specified in building blocks, but in the simulation, e.g. blood/plasma concentration ratio or P (interstitial->intracellular)). Same principle as in case of inherited building blocks: simulation parameters can be inherited between projects. Each inherited simulation parameter description consists of:

  • "Project": Id of the parent project

  • "Simulation": Simulation name within the parent project

  • "Path": Path to the simulation parameter

  • "TargetSimulations": Simulation name(s) within child project

Example

• In all target simulations shown below (DDI Control - xxx, DDI Treatment - xxx), the value of parameter P (interstitial->intracellular) defined in Neighborhoods|Duodenum_int_Duodenum_cell|Midazolam will be set to the value of the same parameter in the simulation po 3mg (solution)

Observed data sets

Similar to a project, an observed data set is identified by its Id, which must be unique within a qualification plan.

There are two kinds of observed data set:

1. Observed data set which is included into one project used by the qualification plan.

   This data set can be used in the qualification plan without any further specification. The Id is, in this case, the name of the observed data set as defined in the PK-Sim project.

2. Observed data set which is not included into one project. It must be described in the "ObservedDataSets" section of a qualification plan.

   

   • "Id": (Unique) id of an observed data set

   • "Path": path to an observed data set file. Can be given as remote URL or local file path (s. the Projects section for details).

   • "Type": type of an observed data set. Can be one of:

• "TimeProfile". Corresponding observed data set must have the columns with time values, measurement values and (optionally) error values with units (s. also here). Example

• "PKRatio". Example. Mandatory columns are:

  • an ID number to reference to ('ID')

  • PK-parameter value and its unit (e.g 'AUC Avg' and 'AUC AvgUnit' for AUC; 'CL Avg' and 'CL AvgUnit' for CL etc.)

  • the simulation duration ('t0' ; 'tend' ; 't Unit').

• "DDIRatio". Example. Mandatory columns are:

  • a unique ID number to reference to ('ID')

  • a unique descriptive name (i.e. Author Year) ('Study ID')

  • the corresponding victim drug ('Victim')

  • the corresponding perpetrator drug ('Perpetrator')

  • the route of administration (e.g. PO, IV, ...) of the victim ('Route Victim')

  • the route of administration (e.g. PO, IV, ...) of the perpetrator ('Route Perpetrator')

  • the compartment (i.e. plasma, whole blood, etc.), from which the victim drug PK parameters should be assessed ('Compartment')

  • the dose of the perpetrator ('Dose') and its unit ('Dose Unit')

  • the observed AUC ratio of the victim expressed as AUC with perpetrator divided by AUC without perpetrator ('AUCR Avg')

  • the observed C_max ratio of the victim expressed as C_max with perpetrator divided by C_max without perpetrator ('CmaxR Avg')

  • the time frame of the simulation without perpetrator, from which the simulated AUC and C_max of the victim should be calculated ('t_placebo_0' ; 't_placebo_end')

  • the time frame of the simulation with perpetrator, from which the simulated AUC and C_max should be calculated ('t_treatment_0' ; 't_treatment_end')

  • and the unit of the respective time definitions ('t Unit').

    Note that observed clearance ratios need to be transformed to AUC ratios before (adding a comment in the 'Comment' column is recommended).

Sections

Defines the chapter structure of the report. A section consists of:

• "Id": Unique section-Id. This id is referenced in other parts of the qualification plan to define which dynamic content must be added to the chapter (dynamic content will be added at the end of the chapter)

• "Title": Chapter title

• "Content": Path to the static content file which will be inserted at the beginning of the chapter. Can be given as remote URL or local file path (s. the Projects section for details). Static content files must be written in Markdown format (s. below).

• "Sections": OPTIONAL list of sub-sections. Every sub-section is built in the same way (thus report structure can be defined with an arbitrary chapter depth level).

Markdown

Markdown files are text files in Markdown format (file extension: .md).

Markdown is a way to style text on the web. You control the display of the document; formatting words as bold or italic, adding images, and creating lists are just a few of the things we can do with Markdown. Mostly, Markdown is just regular text with a few non-alphabetic characters thrown in, like # or *.

Good introductions into the markdown format can be found here:

• https://guides.github.com/features/mastering-markdown/

• https://help.github.com/en/articles/basic-writing-and-formatting-syntax

You can use any plain text editor for creating and modification of markdown files. However it is much faster and easier to use a dedicated markdown editor, e.g. Typora (https://www.typora.io/)

Intro

An (optional) introduction can be added to the report. The differences between introduction and sections are:

1. In the generated report, the introduction will be inserted at the very beginning, before the TOC (table of content) and is not part of the TOC.

2. The introduction does not have any Id and it is not possible to assign any dynamic content to it.

The introduction is defined by:

• "Path": Path to the static content file. Can be given as remote URL or local file path (s. the Projects section for details). Static content files must be written in Markdown format.

Inputs

A convenient way to specify which building blocks and/or simulations should be described in the report as well as the section of the report where the descriptions should be located.

Each input entry definition consists of:

• "Project": Id of the project

• "Name": name of the building block or simulation to describe

• "Type": type of the building block/simulation (one of: "Compound", "Event", "Formulation", "Individual", "ObserverSet", "Population", "Protocol", "Simulation")

• "SectionId": Id of the section where the input description will be inserted.

Input description contains all input settings (model- type, calculation methods etc.) and all input parameters that deviate from the default incl. their value origins.

Plots

This section defines the type of plots (and some additional related information like tables and qualification measures) to generate for the report.

• "PlotSettings": OPTIONAL Global plot settings (pictures size, font properties). In addition, every plot can define its local plot settings.

  • If both (global and local) plot settings are defined for some plot: local settings will be used.

  • If neither global nor local plot settings are defined for some plot: program defaults will be used.

• "AxesSettings": OPTIONAL Global axes settings per plot type. In addition, every plot can define its local axes settings.

  • If both (global and local) axes settings are defined for some plot: local settings will be used.

  • If neither global nor local axes settings are defined for some plot: program defaults will be used.

  • Global axes settings can not be defined for "AllPlots"

  "AxesSettings": {
    "ComparisonTimeProfile": [
      {
        "Unit": "h",
        "Dimension": "Time",
        "Type": "X",
        "GridLines": false,
        "Scaling": "Linear"
      },
      {
        "Unit": "ng/ml",
        "Dimension": "Concentration (mass)",
        "Type": "Y",
        "GridLines": false,
        "Scaling": "Log"
      }
    ],

• "AllPlots"; "GOFMergedPlots"; ... : different kinds of plots, explained in detail below.

AllPlots

All plots defined in the PK-Sim project Project under simulation Simulation will be placed into the report using their settings defined in the PK-Sim project. Thus one node from the "AllPlots"-section in the qualification plan will be expanded into N (N>=0) plots in the final report

"AllPlots": [
  {
    "SectionId": 13,
    "Project": "Midazolam",
    "Simulation": "iv 0.001 mg (5 min)"
  },
  {
    "SectionId": 13,
    "Project": "Midazolam",
    "Simulation": "iv 0.05 mg/kg (2 min)"
  },

NOTE: at the moment, only Time Profile Plots (Individual and Population) will be exported.

GOFMergedPlots

"GOFMergedPlots": [
  {
    "SectionId": 14,
    "Title": "Midazolam concentration in plasma/blood",
    "PlotType": "predictedVsObserved|residualsOverTime",
    "Artifacts": ["Plot", "Measure", "GMFE"],
    "Groups": [
      {
        "Caption": "Midazolam iv",
        "Symbol": "Circle",
        "OutputMappings": [
          {
            "Project": "Midazolam",
            "Simulation": "iv 0.001 mg (5 min)",
            "Output": "Organism|PeripheralVenousBlood|Midazolam|Plasma (Peripheral Venous Blood)",
            "ObservedData": "Hohmann 2015 - iv 0.001 mg - Plasma - agg. (n=16)",
            "Color": "#FF0000"
          },
          ...

Two types of plots are supported here:

• Predicted vs. Observed

• Residuals over time

Combines data from several simulations; every simulation data can be displayed in different color/symbol.

• "Title": title of the plot

• "SectionId": Id of the section where the plot (and related artifacts; s. below) will be inserted.

• "PlotType": One of "predictedVsObserved", "residualsOverTime", "predictedVsObserved|residualsOverTime"

  • if "predictedVsObserved|residualsOverTime" was selected: both plots will be generated

• "Artifacts": OPTIONAL must contain a subset of {"Plot", "Measure", "GMFE"}. Defines which artifacts will be generated in the report. If omitted: all artifacts will be generated

  • "Plot": Plot(s) as described above

  • "Measure": Table with the percentage of data points within X-Error fold

    	Number	Ratio [%]
    Points total	456	-
    Points within 1.5-fold	400	87,7
    Points within 2-fold	440	96,5
  • "GMFE": geometric mean fold error

• "Groups": several simulations can be grouped All simulations from the same group have the same symbol in the plot

  • "Caption": group caption for the plot legend

  • "Symbol": one of "Asterisk", "Circle", "Cross", "Diamond", "Point", "Square", "Triangle"

  • "OutputMappings": definition of pairs {Simulated output <=> Observed data set}

    • "Project": Id of the project

    • "Simulation": name of the simulation

    • "Output": path of the simulated output curve of interest. This must be the path internally used by PK-Sim (without the leading simulation name)

      If you are not sure how such a path is defined:

      1. Open project in PK-Sim

      2. From the context menu of the simulation of interest: select "Export simulation structure to file..."

      3. Open exported file with a text editor and look for "OBSERVER"

      4. In the OBSERVER section: look for your output of interest and copy its path without the leading simulation name.

      E.g. in the example below correct output path for the qualification plan would be

      Organism|PeripheralVenousBlood|Theophylline|Blood Cells

      --------------- OBSERVER (Non zero only) ----------
      Observer: Blood Cells
          Path: S1_diss|Organism|PeripheralVenousBlood|Theophylline|Blood Cells

    • "ObservedData": Id of an observed data set (s. Observed data sets for details)

    • "Color": Color in "#RRGGBB" format. There are numerous free tools for color generation, e.g. https://www.w3schools.com/colors/colors_picker.asp

ComparisonTimeProfilePlots

Creates comparison time profile plots similar to Comparison Charts in PK-Sim. In addition, original results may be shifted in time.

"ComparisonTimeProfilePlots": [
  {
    "SectionId": 26,
    "Title": "Ahonen 1995",
    "SimulationDuration": 20,
    "TimeUnit": "h",
    "OutputMappings": [
      {
        "Project": "Itraconazole-Midazolam-DDI",
        "Simulation": "DDI Control - Midazolam - Ahonen 1995",
        "Output": "Organism|PeripheralVenousBlood|Midazolam|Plasma (Peripheral Venous Blood)",
        "ObservedData": "Ahonen 1995 - Midazolam - PO - 7.5 mg - Plasma - agg. (n=12)",
        "StartTime": 0,
        "TimeUnit": "h",
        "Color": "#2166ac",
        "Caption": "Control (without Itraconazole)",
        "Symbol": "Circle"
      },

• "Title": title of the plot

• "SectionId": Id of the section where the plot (and related artifacts; s. below) will be inserted.

• "SimulationDuration" and "TimeUnit": s. below

• "OutputMappings": definition of pairs {Simulated output <=> Observed data set}

  • "Project": Id of the project

  • "Simulation": name of the simulation

  • "Output": path of the simulated output curve of interest. This must be the path internally used by PK-Sim (without the leading simulation name) (s. GOFMergedPlots for details!)

  • "StartTime" and "TimeUnit": Simulated and observed data will be shifted in the plot:

    • curves will be shifted along the time axis so that original "StartTime" corresponds to Time=0 in the plot

    • only the time range [StartTime .. StartTime + SimulationDuration] of the original data will be plotted;

  • "ObservedData": Id of an observed data set (s. Observed data sets for details)

  • "Color": Color in "#RRGGBB" format. (s. GOFMergedPlots for details). Will be used for both simulated output and observed data

  • "Symbol": Symbol (s. GOFMergedPlots for details) - will be used for observed data only

DDIRatioPlots

Creates DDI Ratio plots as described e.g. in Hanke et. al ([106])

Two types of plots are supported here:

• Predicted vs. Observed (generates plots like in the example above)

• Residuals vs. Observed (generates plots Predicted/Observed vs. Observed)

  "DDIRatioPlots": [
    {
      "SectionId": 5,
      "Title": "CYP3A4 DDI",
      "PKParameter": "AUC|CMAX",
      "PlotType": "predictedVsObserved|residualsVsObserved",
      "Artifacts": ["GMFE","Measure","Plot","Table"],
      "Groups": [
        {
          "Caption": "Itra+Mida (Mida iv)",
          "Color": "#FF0000",
          "Symbol": "Square",
          "DDIRatios": [
            {
              "Output": "Organism|PeripheralVenousBlood|Midazolam|Plasma (Peripheral Venous Blood)",
              "ObservedData": "DDI Ratios",
              "ObservedDataRecordId": 378,
              "SimulationControl": {
                "Project": "Itraconazole-Midazolam-DDI",
                "Simulation": "DDI Control - Midazolam - Olkkola 1996 (iv, day 4)",
                "StartTime": 0,
                "EndTime": 9999,
                "TimeUnit": "h"
              },
              "SimulationDDI": {
                "Project": "Itraconazole-Midazolam-DDI",
                "Simulation": "DDI Treatment - Itraconazole/Midazolam - Olkkola 1996",
                "StartTime": 74,
                "EndTime": 122,
                "TimeUnit": "h"
              }
            }
          ]
        },

• "Title": title of the plot

• "SectionId": Id of the section where the plot (and related artifacts; s. below) will be inserted.

• "PKParameter": PK Parameter for which DDI Ratios will be calculated. One of "AUC", "CMAX", "AUC|CMAX"

  • if "AUC|CMAX" was selected: 2 plots will be generated (one for AUC Ratio and one for CMAX Ratio)

• "PlotType": One of one of "predictedVsObserved", "residualsVsObserved", "predictedVsObserved|residualsVsObserved"

  • if "predictedVsObserved|residualsVsObserved" was selected: both plots will be generated for each selected PK-Parameter. Thus selecting this option in combination with "AUC|CMAX" will result in generation of 4 plots in the report:

    • AUC Ratio predicted vs. observed

    • AUC Ratio residuals vs. observed

    • CMAX Ratio predicted vs. observed

    • CMAX Ratio residuals vs. observed

• "Artifacts": OPTIONAL must contain a subset of {"Plot", "Measure", "GMFE", "Table"}. Defines which artifacts will be generated in the report. If omitted: all artifacts will be generated

  • "Plot": Plot(s) as described above

  • "Measure": Table with the percentage of data points within X-Error fold

    -	Number	Ratio [%]
    Points total	456	-
    Points within Guest et. al	400	87,7
    Points within 2-fold	440	96,5
  • "GMFE": geometric mean fold error (s. GOFMergedPlots for details)

  • "Table": creates a table containing quantitative values of all predicted and observed AUC,CMAX and corresponding DDI-Ratios and additional information about Control and DDI simulation (similar to the table described in Hanke et. al)

• "Groups": plotted DDI ratios can be grouped. Each group has its own caption, color and symbol

  • "Caption": plot caption

  • "Color": color in "#RRGGBB" format. (s. GOFMergedPlots for details).

  • "Symbol": group symbol (s. GOFMergedPlots for details)

  • "DDIRatios": list of DDI ratios belonging to the group. Each DDI Ratio is defined by:

    • "Output": path of the simulated output curve for which DDI ratio of interest will be calculated. This must be the path internally used by PK-Sim (without the leading simulation name) (s. GOFMergedPlots for details!).

    • "ObservedData": Id of an observed data set (s. Observed data sets for details)

    • "ObservedDataRecordId": Id of the data record (line) within the given observed data set. (corresponds to the Id-column of the data set)

    • "SimulationControl": description of the Control/Placebo simulation, given by:

      • "Project": Id of the project

      • "Simulation": name of the simulation

      • "StartTime", "EndTime" and "TimeUnit": PK-Parameter(s) of interest will be calculated in the time range [StartTime .. EndTime]

    • "SimulationDDI": description of the DDI simulation, given in the same way as Control simulation

PKRatioPlots

Creates plots of predicted/observed ratios for PK parameters of interest

"PKRatioPlots": [
  {
    "Title": "Overall predictivity of the PBPK models. Open circles represent...",
    "SectionId": 2,
    "PKParameter": "AUC|CL",
    "Artifacts": ["GMFE", "Measure", "Plot", "Table"],
    "Groups": [
      {
        "Caption": "Caption",
        "Color": "#000000",
        "Symbol": "Circle",
        "PKRatios": [
          {
            "Project": "Sufentanil",
            "Simulation": "Davis 1987 15.5months",
            "Output": "Organism|ArterialBlood|Plasma|Sufentanil|Concentration in container",
            "ObservedData": "PK-Parameter",
            "ObservedDataRecordId": 5130
          }

• "Title": title of the plot

• "SectionId": Id of the section where the plot (and related artifacts; s. below) will be inserted.

• "PKParameter": PK Parameter for which DDI Ratios will be calculated. One of "AUC", "CL", "AUC|CL"

  • if "AUC|CL" was selected: 2 plots will be generated (one for AUC and one for Clearance)

• "Artifacts": OPTIONAL must contain a subset of {"Plot", "Measure", "GMFE", "Table"}. Defines which artifacts will be generated in the report. If omitted: all artifacts will be generated

  • "Plot": Plot(s) as described above

  • "Measure": Table with the percentage of data points within X-Error fold (s. GOFMergedPlots for details)

  • "GMFE": geometric mean fold error (s. GOFMergedPlots for details)

  • "Table": creates a table containing quantitative values of all predicted and observed AUC and/or CL and corresponding PK-Ratios, e.g.

    Study ID	Age [y]	BodyWeight [kg]	Predicted AUC [µmol*h/l]	Observed AUC [µmol*h/l]	Pred/Obs AUC Ratio
    Larson 2013	15	56	1.4103	15.7	1.4103
    Larson 2013	9	29.4	0.85609	18	0.85609
    Nachmann 2013	15.2	56.2267	1.0989	10.2	1.0989
    Nachmann 2013	10	31.5	1.063	13.4	1.063
    Rizk 2015	1.25	8.5	1.7188	19.8	1.7188
• "Groups": plotted PK ratios can be grouped. Each group has its own caption, color and symbol

  • "Caption": plot caption

  • "Color": color in "#RRGGBB" format. (s. GOFMergedPlots for details).

  • "Symbol": group symbol (s. GOFMergedPlots for details)

  • "PKRatios": list of PK ratios belonging to the group. Each PK Ratio is defined by:

    • "Project": Id of the project

    • "Simulation": name of the simulation

    • "Output": path of the simulated output curve for which DDI ratio of interest will be calculated. This must be the path internally used by PK-Sim (without the leading simulation name) (s. GOFMergedPlots for details!).

    • "ObservedData": Id of an observed data set (s. Observed data sets for details)

    • "ObservedDataRecordId": Id of the data record (line) within the given observed data set. (corresponds to the Id-column of the data set)

How generated artifacts are combined into a report

All static and dynamic elements described in a qualification plan are compiled into a report in the following order:

1. Intro (if defined in the qualification plan)

2. Table of Contents (is generated automatically)

3. (Top level) sections in order of their appearance in the qualification plan. Per section:

   3.1 Static content of the section

   3.2 For all inputs with SectionId = Id of the current section: generated input descriptions in order of appearance in the qualification plan

   3.3 For all plots with SectionId = Id of the current section: generated plots (and related artifacts) in order of appearance in the qualification plan

   3.4 Subsections of the current section (if any) in order of appearance in the qualification plan. Per subsection ... (s. 3.1..3.4)

Creating a (re-)qualification plan part II: Tools

1. Install VSCode (Visual Studio Code).

   • Download User-Installer or System-Installer from https://code.visualstudio.com/Download

2) If you are behind a firewall: configure firewall proxy

• Start VSCode

• Go to File►Preferences►Settings

• Then go to User Settings►Application►Proxy

• Enter your Firewall-Proxy

1. Install the snippets file

   • Download Snippets for Visual Studio Code (qualification.code-snippets) from https://github.com/Open-Systems-Pharmacology/QualificationPlan/releases/latest

   • Copy this file to C:\Users\<USERID>\AppData\Roaming\Code\User\snippets

     • In case you are using portable version of VSCode: copy the snippets file to <VSCode_InstallDir>\data\user-data\User\snippets (create the folder if it does not exist)

   • Restart VSCode

4) Create/Edit a qualification plan

• Create a new empty file and save it as .json (unless the file was saved as json, snippets will not work)

• For every element of a qualification plan (project, plot, section, input) there is a predefined code snippet, which will create a skeleton of this element.

• To use a snippet, type its shortcut and press ENTER

• The full list of qualification plan snippets is given in the table below. Parent node column describes at which places of a qualification plan a snippet can be used.

• The first snippet to be used is always bs(bootstrap) - this will create a skeleton of a qualification plan

• All other snippets are defined as abbreviations of "Add XYZ".

• Full snippets list

Shortcut	Description	Parent node
bs	Creates the skeleton for a qualification plan	•ROOT
ap	Add a project reference	•Projects
abb	Adds a building block reference. Useful to replace a building block in a project	•Projects{i}/BuildingBlocks
asp	Adds a simulation parameter reference. Useful to replace a parameter in a simulation with a parameter from another simulation	•Projects{i}/SimulationParameters
aod	Adds an observed data reference. Only for external observed data sources	•ObservedDataSets
ai	Adds an input (reference to a building block or simulation in a given project)	•Inputs
aintro	Adds an introduction chapter	•Intro
as	Adds a section	•Sections •Sections{i}/Sections …
aps	Adds the default plot settings configuration (global or local)	==== GLOBAL ==== •Plots ==== LOCAL ==== •GOFMergedPlots{i} •ComparisonTimeProfilePlots{i} •DDIRatioPlots{i} •PKRatioPlots{i}
aas	Adds the global axes settings skeleton	•Plots/AxesSettings
axy	Adds axes X and Y settings content (global or local)	==== GLOBAL ==== •Plots/AxesSettings/GOFMergedPlotsPredictedVsObserved •Plots/AxesSettings/GOFMergedPlotsResidualsOverTime •Plots/AxesSettings/ComparisonTimeProfile •Plots/AxesSettings/DDIRatioPlotsPredictedVsObserved •Plots/AxesSettings/DDIRatioPlotsResidualsVsObserved •Plots/AxesSettings/PKRatioPlots ==== LOCAL ==== •Plots/GOFMergedPlots{i}/AxesPredictedVsObserved •Plots/GOFMergedPlots{i}/AxesResidualsOverTime •Plots/ComparisonTimeProfilePlots{i}/Axes •Plots/DDIRatioPlots{i}/AxesPredictedVsObserved •Plots/DDIRatioPlots{i}/AxesResidualsVsObserved •Plots/PKRatioPlots{i}/Axes
aap	Adds an all plot entry for a simulation	•Plots/AllPlots
agof	Adds a GOFMergedPlot entry	•Plots/GOFMergedPlots
agofg	Adds a GOFMergedPlot group entry	•Plots/GOFMergedPlots{i}/Groups
agofo	Adds a GOFMergedPlot OutputMapping entry (to be used within a group)	•Plots/GOFMergedPlots{i}/Groups{j}/OutputMappings
actp	Adds a ComparisonTimeProfile entry	•Plots/ComparisonTimeProfilePlots
actpo	Adds a ComparisonTimeProfile OutputMapping entry	•Plots/ComparisonTimeProfilePlots{i}/OutputMappings
addir	Adds a DDIRatioPlot entry	•Plots/DDIRatioPlots
addirg	Adds a DDIRatioPlot group entry	•Plots/DDIRatioPlots{i}/Groups
addirr	Adds a DDIRatioPlot ratio entry (to be used within a group)	•Plots/DDIRatioPlots{i}/Groups{j}/DDIRatios
apkr	Adds a PKRatioPlot entry	•Plots/PKRatioPlots
apkrg	Adds a PKRatioPlot group entry	•Plots/PKRatioPlots{i}/Groups
apkrr	Adds a PKRatioPlot ratio entry (to be used within a PKRatioPlot)	•Plots/PKRatioPlots{i}/Groups{j}/PKRatios

• If you do not remember the shortcut of a snippet:

  _ either start typing: the list of all snippets starting with this shortcut will be shown via Intellisense

  _ or press CTRL+SPACE: the list of ALL snippets will be shown. Then just navigate to the right snippet and select it

• After you inserted a skeleton via snippet: fill all the placeholders with correct information. Just start typing (don't click with the mouse!) into the first entry; once finished - press TAB to switch to the next input

• If a value to be entered is an enumeration: click between double quotes and press CTRL+SPACE, then select from the list

• If a Dimension/Unit pair has to be defined: select the dimension first (CTRL+SPACE), AFTER that select the unit

• Every time when a new element of a qualification plan was entered via snippet (or manually) and filled out: immediately check errors and warnings and correct them as soon as possible.

• When adding a new element of NON-EMPTY array, do not forget a comma before or after inserted element. (Before when inserted as last element, after otherwise).

• Some helpful links for editing json files with VSCode:

  • Getting started with VSCode: https://code.visualstudio.com/docs/getstarted/introvideos

  • Editing JSON with VSCode: https://code.visualstudio.com/docs/languages/json

  • Extending/Modifying Snippets: https://code.visualstudio.com/docs/editor/userdefinedsnippets

Processing a (re-)qualification plan

Tools

Creation of a qualification report from a qualification plan requires installation of additional tools, which are not part of the OSP Suite setup. All required tools can be downloaded from

https://github.com/Open-Systems-Pharmacology/QualificationPlan/releases/latest

• QualificationRunner: download qualificationrunner-portable-setup_X.Y.Z.zip and unzip it into any folder on your hard disc.

• Reporting Engine. There are two installation options:

  1. In case you have a Matlab license: install as a source code. For this: download Reporting.Engine.X.Y.Z.zip and unzip it into any folder on your hard disc.

  2. In case you have no Matlab license: install as compiled library.

     1. Prerequisite: download and install free Matlab Compiler Runtime version 2017b

        http://ssd.mathworks.com/supportfiles/downloads/R2017b/deployment_files/R2017b/installers/win64/MCR_R2017b_win64_installer.exe

     2. Download Reporting.Engine.Compiled.X.Y.Z.zip and unzip it into any folder on your hard disc.

     3. Unzip folder contains the file CreateQualificationReport.bat. Adjust it (s. comments in the file)

• Markdown Joiner: Download markdown-joiner.zip and unzip it into any folder on your hard disc.

Creating a report in Markdown format

A good starting point is https://github.com/Open-Systems-Pharmacology/Evaluation-plan-template.

Download this repository locally and adjust Workflow.m in the subfolder Evaluation (s. comments in the file).

Execute Workflow.m.

• If you have a Matlab license: start Matlab and execute Workflow.m

• If you have no Matlab license:

  1. Start command prompt (cmd). In the command prompt:

     1. Switch to the folder where Reporting.Engine.Compiled.X.Y.Z.zip was unzipped.

     2. Execute CreateQualificationReport.bat "". E.g.

        CreateQualificationReport.bat "C:\Evaluation-plan-template\Evaluation\Workflow.m"

This will create a report in markdown format in the Evaluation/report subfolder.

For any new qualification report: just create a new copy of Workflow.m and adjust it for the new report.

Converting Markdown report to pdf.

Different (commercial and free) markdown to pdf converters are available. We recommend to use Typora (https://www.typora.io/) for this task.