ACoP9-Workshop: Using R and Shiny to support efficient PK/PD and QSP modeling, parameter estimation, simulation, and reporting in Microsoft Word

ACoP9 Meeting, October 6th Saturday, Loews Coronado Bay Resort near San Diego, CA, US

Workshop Description

As R itself is pretty powerful, it does not come right out-of-the-box with support for the efficient conduct of pharmacometric or QSP modeling and simulation activities. Available R packages on CRAN support some aspects but typically lack in user-friendliness, stability, and functionality. For this reason, the main part of this workshop will be conducted using the IntiQuan R Tools (IQRtools), which is a comprehensive R package available for download and free use for all workshop participants.

  • Morning session
    • PK, PBPK, PKPD, and QSP modeling in Shiny
    • Behind the “Shiny scene”: IQR Tools
      • How to define and simulate models
      • Script based NLME parameter estimation in NONMEM, MONOLIX & NLMIXR
      • Generation of report-ready model diagnostics, parameter estimates tables, and model development tables
  • Afternoon session
    • Robust parameter estimation in QSP based on sensitivity equations and multi-start optimization
    • Hands-on examples
    • Optimal experiment design via seamless PopED interface
    • Analyzing QSP models, informing modeling decisions, using profile likelihood
    • Reporting to Microsoft Word
    • Concluding Remarks

After the workshop the participants will be able to write PK, PK/PD and more mechanistic QSP models, simulate them, and perform parameter estimation in supported NLME tools (such as NONMEM, MONOLIX, and NLMIXR), and using a QSP type of parameter estimation method. Furthermore, participants will be able to generate standard model diagnostics and export all results to Microsoft Word.

The workshop is designed as a hands-on tutorial, allowing participants to work through all examples under guidance of the presenters. All material (slides, example models, R code, required R libraries, and additional software) will be shared with the participants. In case a participant would like to have hands-on experience during the workshop, a laptop with an R installation should be brought to the workshop. Some examples will use NONMEM and/or MONOLIX at the backend for parameter estimation, so it would be good if either of these tools were installed already. In the case that neither is available, the examples can be run with the open-source and freely available NLMIXR R package instead. No prior knowledge of either NONMEM, MONOLIX, or NLMIXR is required. Basic knowledge of NLME modeling and R is of advantage.

Program*

October 6th Saturday
Time Topic Presenter
8:00-8:15 Welcome and Introduction Henning Schmidt
8:15-10:00
  • PK, PBPK, PKPD, and QSP modeling in Shiny
Henning Schmidt
10:00-10:15 Coffee break
10:15-12:00
  • Behind the “Shiny scene”: IQR Tools
    • How to define and simulate models
    • Script based NLME parameter estimation in NONMEM, MONOLIX & NLMIXR
    • Generation of report-ready model diagnostics, parameter estimates tables, and model development tables
    • Reporting to Microsoft Word (quick intro – more details in second IntiQuan Workshop)
Henning Schmidt
12:00-13:00 Lunch
13:00-14:00
  • Robust parameter estimation in QSP based on sensitivity equations and multi-start optimization
Daniel Kaschek
14:00-15:00
  • Hands-on examples
Henning Schmidt & All
15:00-15:15 Coffee break
15:15-16:45
  • Optimal experiment design via seamless PopED interface
  • Analyzing QSP models, informing modeling decisions, using profile likelihood
  • Reporting to Microsoft Word
Daniel Kaschek
16:45-17:00 Concluding remarks

Henning Schmidt

Daniel Kaschek

* Exact agenda and content is subject to small changes to adapt to new developments and profile of the audience

Presenters

  Daniel Kaschek, PhD, IntiQuan GmbH

Daniel Kaschek is an experienced modeler working in the field of Quantitative Systems Pharmacology and Systems Biology since 2008. Over the past years, he has actively worked on the developed of novel mathematical and statistical approaches to data analysis and parameter estimation in ordinary differential equations with applications in Quantitative Systems Pharmacology and Systems Biology. He is an enthusiast R package developer, authoring several open-source packages on Data pre-processing and dynamic modeling.

  Henning Schmidt, PhD, IntiQuan GmbH

Engineer by training with a long background in modeling of a wide range of general linear and non-linear systems (automotive & process industry, biological, physiological, and pharmacological systems), using both modeling and simulation results as a basis for impactful quantitative support of decision making in large and small, early and late phase drug development projects. Developer of several open source tools for modeling and simulation across Pharmacometrics, Systems Pharmacology, and Systems Biology.

Registration

Workshop location: The workshop will be held at the Loews Coronado Bay Resort near San Diego, CA, US, which is the venue for the ACoP9 meeting.

Workshop fee: Regular fee 600 USD. This includes workshop material, coffee breaks, and lunch. Students/Academia/Government/Non-profit employees may enroll at a fee of 300 USD.

Registration: This is a pre-meeting workshop that requires registering for the ACoP 2018 meeting. Please register ASAP in view of the limited course capacity.

Contact: Henning Schmidt, IntiQuan GmbH, Spalenring 150, 4055 Basel, Switzerland; Phone: +41 76 603 28 06; Email: henning.schmidt@intiquan.com.