- Home
ACoP 2024 Workshops & Tutorials
November 9, 20248:00 am - 5:00 pm Scientific Modeling Augmented by Machine-Learning with DeepPumas, PumasAI (Day 1)Description: DeepPumas is a powerful tool that bridges the gap between mechanistic, statistical, and machine-learning modeling. This hands-on workshop will introduce participants to the core concepts behind DeepPumas and its potential to transform decision-making in the healthcare space. By seamlessly embedding machine learning methodology into nonlinear mixed-effect models, DeepPumas enables the simultaneous utilization of domain-specific knowledge and available data for model identification. In this workshop, we will learn about NeuralODEs, universal differential equations, and scientific machine learning and how these can be leveraged to discover unknown dynamics driving disease progression and treatment response. We will then see how DeepNLME expands these concepts to include random effects which lets us discover individualizable models that can tell us something about not only the measurements in our data but about the patients whose outcomes we care about. Furthermore, we will learn how to leverage machine learning to identify prognostic factors from complex data. After gaining theoretical understanding and practical experience in using these new technologies, we will explore different use cases to understand how DeepPumas is poised to transform the role of data and modeling in healthcare.
Presenters: Niklas Korsbo, Mohamed Tarek, Vijay Ivaturi, Andreas Noack
Efficient reproducible Bayesian population PK modeling with NONMEM and Stan/Torsten, Metrum Research Group (Day 1)Description: This workshop provides a guided hands-on experience in performing efficient reproducible and traceable Bayesian PPK analyses. We will walk through both NONMEM-based and Stan-based workflows. Such workflows are facilitated by the use of MeRGE (Metrum Research Group Ecosystem) open-source tools that have been extended to support fully Bayesian analyses with either NONMEM or Stan. Sometimes you may want more flexibility than offered by NONMEM. We illustrate that by implementing a PPK model employing shrinkage priors for covariate effects using Stan/Torsten.
Presenters: Bill Gillespie, Tim Waterhouse, Curtis Johnston, Seth Green
Hands-On Introduction to InSilicoTrials' Cloud-Based Clinical Trial Simulator, InsilicoTrialsDescription: In this hands-on workshop we will provide an introduction to our cloud-based trial simulator. This technology is fully browser-based, and provides users with a powerful yet intuitive and user-friendly interface for performing clinical trial simulations. Furthermore, if the underlying models were developed using NONMEM, simulations may be performed without any modification of the NONMEM code.
Learning Objectives: Key principles of clinical trial simulation (CTS), Notable case studies illustrating the application and value of CTS, How to perform CTS using InSilicoTrials’ trial simulator
Presenters: Mark Lovern, Daniel Roeshammar, Nathan Teuscher
New and Advanced Features of NONMEM 7.5 Workshop, ICON Clinical Research LLCDescription: Advanced NONMEM 7.5 course will be presented by ICON at ACOP 2024. This in-person workshop covers the description and use of features in NONMEM 7. Workshop attendees will be instructed how to specify gradient precision and how to use the FAST algorithm (new in NM 7.4) for FOCE, how to use the Monte Carlo importance sampling, stochastic approximation expectation-maximization methods, and full Bayesian methods such as Gibbs sampling and Hamiltonian no-U turn sampling (new in NM 7.4). Parallel computing and dynamic memory allocation for efficient memory usage will be described, symbolic references to thetas, etas, and sigmas, priors to sigmas, MonteCarlo search algorithms to improve FOCE estimation, built-in individual weighted residuals, bootstrap tools for simulation, and automatic stabilization against numerical exceptions. Learn to use abbreviated code features for easier modeling of inter-occasion variability, modeling additional mixed effects levels for grouping individuals, such as inter-clinical site variability, and using the DO loop feature in abbreviated code, useful for handling multiple bolus doses in models that use the analytical absorption function for multiple transit compartments. Optimal clinical design and evaluation tool is available, as well as delay differential equation solvers.
Presenters: Robert Bauer, Brian Sadler
1:00 – 5:00 pmPractical Tools for Model Calibration with SimBiology, MathWorksDescription: Practical Tools for Model Calibration with SimBiology Mathematical modeling such as (PB)PK/PD and QSP is becoming increasingly important to inform decision making in various stages of the drug development process. These models often take the form of ordinary differential equations and include numerous parameters that need to be estimated with experimental data. The process of model calibration presents many challenges such as the difficulty to assess parameter identifiability, scarcity of high-quality experimental data, and highly nonlinear objective functions with multiple local optima. In this workshop, we will cover a workflow for model calibration in SimBiology which encompasses methods for Monte Carlo sampling, sensitivity analysis, data fitting, and parameter identifiability analysis. These techniques, when used in sequence, establish a comprehensive approach for effective model calibration. We will provide a theoretical background for each of these steps and offer step-by-step hands-on exercises to build up a full model calibration workflow derived from a real-world example. By the end of this workshop participants will have learned: - How to generate and interpret scatterplots to identify non-influential parameters of a model. - How to perform global sensitivity analysis (GSA) with SimBiology and leverage the results to select a subset of model parameters for estimation, while understanding the distinctions among the various GSA methods implemented. - The difference between local and global optimization methods and the reasons why global solvers are often preferred for fitting data. - How to use the Profile likelihood method to detect parameter non-identifiability.
Presenters: Fulden Buyukozturk, Marco Avila Ponce De Leon
November 10, 20248:00 am – 5:00 pm Scientific Modeling Augmented by Machine-Learning with DeepPumas, PumasAI (Day 2) Efficient reproducible Bayesian population PK modeling with NONMEM and Stan/Torsten, Metrum Research Group (Day 2)Biologics (ADC) whole-body PBPK/QSP in PK-Sim and MoBi, ESQlabs GmbHDescription: The development of large-molecule drugs has become increasingly complicated over the last decades due to the continuous addition of new modalities and mechanisms of action. The crucial role of biodistribution into multiple tissues for target binding, catabolism, and FcRn-mediated recycling throughout the body makes whole-body PBPK modeling an attractive option for model-informed drug development in this area. These models are especially useful if they can be extended to include target-mediated drug disposition (TMDD) in multiple organs where the target is expressed, and if pharmacodynamic elements can be added to obtain integrated PBPK-QSP models. ESQlabs will provide hands-on training on the development of these models using PK-Sim® and MoBi® (OSP-Suite). The workshop will cover general introductions to applying this freely available open-source software suite and will provide state-of-the-art example models and development workflows. In-depth exercises will cover PBPK and PD modeling for monoclonal antibodies, antibody-drug-conjugates (ADCs) for tumor growth inhibition, FcRn inhibitors, and bispecific immune cell engagers.
Presenters: Wilbert De Witte, Alexander Kulesza
Modeling Delays in Pharmacokinetics and Pharmacodynamics using Phoenix, ISoP MCS SIGDescription: Mathematical and Computational Sciences Special Interest Group of International Society of Pharmacometrics will present a one-day workshop on Modeling Delays in Pharmacokinetics and Pharmacodynamics using Phoenix. The course will provide an overview of biological systems exhibiting delays, concepts of lifespan driven pharmacodynamic response, modeling of cell maturation, transduction delays, and nonlinear mixed effect lifespan models. Delay differential equations (DDEs) and distributed delay differential equations (DDDEs) will be introduced and implemented in Phoenix. The course will consist of both lectures and hands-on computer exercises. Source codes for DDE and DDDE based PK/PD models and data will be provided to the participants.
Learning Objectives: Learn biology and pharmacology underlying in PK/PD systems, Introduce mathematical and computational basics of DDEs and DDDEs, Gain hands-on experience with implementing PK/PD models in Phoenix
Presenters: Wojciech Krzyzanski, Shuhua, Hu
Simcyp Designer PBPK-QSP Focused Workshop, CertaraDescription: This year, Certara UK is running a one-day Simcyp hands-on workshop on developing population-PBPK driven PD/QSP (quantitative systems pharmacology) models within its new integrated platform "Simcyp Designer". The course reviews the fundamentals of PBPK and QSP modelling using the Simcyp Simulator and builds rapidly to specialised applications where you will modify existing PBPK models with your own PD/QSP extensions. We will also examine the propagation of population variability to mechanistic models of therapeutic effects dynamically interacting with drug distribution models, as well as analysing population variability in clinical data.
The key topics covered are:
*Latest advances of PBPK modelling using the Simcyp Simulator,
*Generation of realistic population variability using correlated Monte Carlo sampling, covariate modeling withing the Simcyp Simulator,
*Replacing or modifying a tissular compartment within the Simcyp Simulator with a user-defined model,
*Creating PD/QSP models using the Simcyp Designer's graphical editor,
*Incorporating physiological covariates within a user-defined mechanistic effect model,
*Adding target mediated drug disposition (TMDD) processes to tissue compartments, •Extending PBPK models for therapeutic proteins,
*Optimising PBPK models using data across different clinical trials simultaneously.
Presenters: Abdallah Derbalah, Felix Stader
November 14, 20248:00 am – 12:00 pm Tutorial: Model-Based Meta-Analysis: towards more precisely predicted clinical scenariosDescription: In recent times there has been a growing interest in reutilizing clinical data available in the public domain to efficiently design clinical trials with new molecular entities, while benchmarking them against the competitor landscape. Model-based meta-analysis (MBMA) is a method to integrate data from multiple studies using mathematical models to quantitatively describe the effect of treatment, time, and patient population characteristics on the trial outcomes. In response to the increasing attention to this methodology, and because this topic offers a unique opportunity for the collaboration between Pharmacometricians and Biostatisticians, the ISoP-ASA Statistics and Pharmacometrics Special Interest Group (SxP SIG) sponsored the creation of the MBMA Special Interest SubGroup (MBMA SubSIG). Since then, the core team, composed by members from different companies and organizations, has pursued its objective of continuing to raise awareness and interest, and to support the education of others in MBMA, among the subSIG members and across the broader audience.
In the last 15 years while MBMA has found application in different indications such as immunology, oncology and infectious diseases, among the others, the attention to the methodology has also increased in the attempt to improve the interpretation of the external data and the precision in predicting the internal outcomes. In this session in addition to an introduction on the state-of-the-art, a closer look will be given to some methodological aspects, like comparing relative, vs. absolute outcome analyses, and handling covariate and variance components. The tutorial will conclude with the perspective from one regulatory authority.
Upon completion, the audience will learn general concepts about Model-Based Meta-Analysis (MBMA), including how to build disease progression models from longitudinal clinical aggregate-level data as well as how to perform landmark analyses. Then the focus will be concentrated on specific methodological aspect like the implications of using relative vs. absolute outcomes. Ways of handling missing covariates as well as variance components will be illustrated. Applications of these methodologies will be shown across selected disease areas and coding examples will be also presented in some instances.
Chairs: Phyllis Chan, Monica Simeoni
Presenters: Monica Simeoni, Matthew Zierhut, Rana Reich, Abhinav Kurumaddali, Elyes Dahmane
Tutorial: Using Past Models to Bridge to Open Models and Open Science using nlmixr2Description: Open science is a movement to make science available to all levels of society. The science in much of population-based pharmacometrics in the past has been focused on developing nonlinear mixed effects models in proprietary tools like NONMEM and Monolix. This makes it necessary to have licenses of whatever tool is being used to be able to explore models with your data. Special populations in low to middle income countries may not have access to these tools which makes analysis of additional clinical data in these regions more challenging. This tutorial discusses automated methods to import NONMEM and Monolix into the open-source framework rxode2 and nlmixr2 using packages like nonmem2rx, monolix2rx and babelmixr2. Once converted, you can use these models to make patient-based adaptive dosing decisions, simulate other dosing scenarios and even use the model to analyze new data and explore any regional differences in drug effect.
Upon completion, participants
1. learn the basics of how a nlmixr2/rxode2 model is written and can write a model themselves.
2. Upon completion, participants will know how to import NONMEM and Monolix models into the nlmixr2/rxode2 model function using nonmem2rx and monolix2rx
3. Upon completion, participants will know how to simulate new dosing scenarios using rxode2
4. Upon completion, participants will know how to (re-)estimate new data using nlmixr2
5. Upon completion, participants learn how to individualize dosing clinically using posologyr
Chair: Matthew Fidler
Presenters: William Denney, Mirjam Trame, Justin Wilkins
9:00 am – 12:00 pmTrainee Tutorial: Minimal PBPK Modeling: Tutorial and ApplicationsIntroduction –Carter Cao, Ph.D. University of North Carolina Eshelman School of Pharmacy
Small Molecule Minimal PBPK Models and Classic Hepatic Clearance Concept – Xiaonan Li, Ph.D. Takeda Pharmaceuticals
Large Molecule Minimal PBPK Models and FcRn Interactions – Dongfen Yuan, Ph.D. Janssen
1:00 – 5:00 pm(
R)TTE analysis: MIDD applications, concept, methodology and NONMEM hands-on, Pharmetheus (Day 1)Description: This 1.5 days workshop’s objective is to provide a good understanding of time to event (TTE) and repeated TTE (RTTE) modeling in NONMEM, in theory and practice, by listening to and discussing with leaders from both industry and academia. Target audiences are students and pharmacometricians with no, or limited, experience of TTE/RTTE modeling.
Day1 (free of charge):
The motivation for running survival analyses from a model-informed drug development perspective with key speakers from academia and industry.
Key speakers: Lena Friberg, Nastya Kassir, France Mentré
Day 2:
An introduction to general concepts and theory, typical TTE/RTTE analysis followed by data formatting and key aspects of graphical exploration of TTE/RTTE data,
Detailed insights to NONMEM code useful for the model development, such as standard and alternative function parameterization, implementation of exposure-response relationships and estimation of covariate effects,
An overview of efficient simulation frameworks, with a focus on NONMEM coding and derived metrics to graphically communicate on,
Several hands-on sessions to help the attendees appreciate specificities of TTE/RTTE models in terms of evaluation and diagnostics.
Workshop instructors: Jurgen Langenhorst, Felicien Le Louedec, Joakim Nyberg, Qing Xi Ooi
After the workshop the attendees will be able to understand, conduct and communicate TTE and RTTE modeling analyses.
For more information visit Events under Pharmetheus webpage.
LAP&P course: An interactive translational pharmacometric case study for an antibody exhibiting TMDD, LAP&P Consultants BV (Day 1)Description: In this 1.5-day course we will specifically discuss the importance of understanding the pharmacology of monoclonal antibodies (mAbs) to make the right decisions in target-mediated drug disposition (TMDD) model development. If you want to learn more on 1) the relevance of mechanistic modelling for mAbs, 2) the considerations with regards to scaling the model for first in human (FIH) approaches, and 3) what TMDD models are, what approximations may be applicable and how to use these models for (non-) clinical data. After a short introduction, we will directly dive into a hands-on session, in which the participants work in small groups on TMDD modelling challenges using a nlmixr2 within a shinyMixR workflow.
Presenters: Tamara van Steeg, Sven Hoefman, Richard Hooijmaijers
November 15, 20248:00 am – 5:00 pm(R)TTE analysis: MIDD applications, concept, methodology and NONMEM hands-on, Pharmetheus (Day 2)Day 2
Description: This 1.5 days workshop’s objective is to provide a good understanding of time to event (TTE) and repeated TTE (RTTE) modeling in NONMEM, in theory and practice, by listening to and discussing with leaders from both industry and academia. Target audiences are students and pharmacometricians with no, or limited, experience of TTE/RTTE modeling.
Day1 (free of charge):
The motivation for running survival analyses from a model-informed drug development perspective with key speakers from academia and industry.
Key speakers: Lena Friberg, Nastya Kassir, France Mentré
Day 2:
An introduction to general concepts and theory, typical TTE/RTTE analysis followed by data formatting and key aspects of graphical exploration of TTE/RTTE data,
Detailed insights to NONMEM code useful for the model development, such as standard and alternative function parameterization, implementation of exposure-response relationships and estimation of covariate effects,
An overview of efficient simulation frameworks, with a focus on NONMEM coding and derived metrics to graphically communicate on,
Several hands-on sessions to help the attendees appreciate specificities of TTE/RTTE models in terms of evaluation and diagnostics.
Workshop instructors: Jurgen Langenhorst, Felicien Le Louedec, Joakim Nyberg, Qing Xi Ooi
After the workshop the attendees will be able to understand, conduct and communicate TTE and RTTE modeling analyses.
For more information visit Events under Pharmetheus webpage.
LAP&P course; An interactive translational pharmacometric case study for an antibody exhibiting TMDD, LAP&P Consultants BV (Day 2)Description: In this 1.5-day course we will specifically discuss the importance of understanding the pharmacology of monoclonal antibodies (mAbs) to make the right decisions in target-mediated drug disposition (TMDD) model development. If you want to learn more on 1) the relevance of mechanistic modelling for mAbs, 2) the considerations with regards to scaling the model for first in human (FIH) approaches, and 3) what TMDD models are, what approximations may be applicable and how to use these models for (non-) clinical data. After a short introduction, we will directly dive into a hands-on session, in which the participants work in small groups on TMDD modelling challenges using a nlmixr2 within a shinyMixR workflow.
Presenters: Tamara van Steeg, Sven Hoefman, Richard Hooijmaijers
A Practitioner's Guide to Identifiability Analysis and Optimal Design of Experiments in Pharmacometrics using Pumas, Pumas AIDescription: Understanding when a statistical model's parameters are identifiable or otherwise, given a specific experiment design, prior to collecting data, is a critical step that is often over-looked. Given that our ability to estimate a model's parameter (aka its practical identifiability) is dependent on the experiment design, one may want to design experiments intelligently to: 1) make the model's parameters that we care about identifiable, and 2) maximise the precision by which we can estimate those parameters. In many cases when designing an experiment, we already know the model we intend to use for analysis. Knowing the analysis model allows us to optimally design our experiment to maximise the expected amount of information we can learn about the parameters that we care about, e.g. to show efficacy and get a drug approved.
Learning Objectives: Understand the math behind the optimal design of experiments, use Pumas to prove a model’s local identifiability or otherwise, Use Pumas to choose the optimal sample times in an experiment to maximise information gain subject to logistic constraints, e.g. a clinic’s operating hours, Use simulation to verify the quality of your chosen experiment design.
Presenters: Mohamed Tarek, Vijay Ivaturi
PBBM - Formulation development and virtual Bioequivalence, ESQlabs GmBHDescription: Model-informed drug development is extensively used in decision-making during drug development. Physiologically based biopharmaceutics modeling (PBBM) is increasingly explored to inform oral drug product development, manufacturing changes, and controls.
Learning Objectives: Delve into the world of mechanistic oral absorption modeling with PK-Sim®, unlocking a deeper understanding of absorption mechanisms; Learn about the in vitro-in vivo correlation (IVIVC) concept and its significance in pharmaceutical development and regulatory applications, from streamlining formulation optimization to expediting regulatory approval processes; Explore ways to establish the IVIVC through the OSP Suite; Learn about virtual bioequivalence (VBE) and dissolution safe space and how these strategies are reshaping traditional approaches in the drug development space; Immerse yourself in interactive exercises, gaining experience in VBE analyses with the OSP suite, empowering you to navigate the complexities of pharmaceutical research.
Presenters: Mariana Guimarães, Marco Siccardi, Jeremy Perrier
