Advanced Machine Learning and NLME Modeling in Drug Development

This five-day workshop is designed to deepen your understanding of non-linear mixed-effect (NLME) modeling and its integration with advanced machine learning techniques in pharmacometrics. Participants will explore the capabilities of Pumas software for drug development and individualized treatment modeling through DeepPumas, learning both theoretical concepts and practical applications. The workshop will culminate in a hands-on case study where participants apply what they’ve learned to real-world data.

Instructors: Niklas Korsbo, PhD and Mohamed Tarek, PhD

Objectives:

1. Develop a foundational understanding of non-linear mixed-effect models and their applications in pharmacometrics using Pumas.
2. Explore the current use of machine learning in drug development and pharmacometrics through case studies and practical exercises.
3. Dive into advanced topics in DeepPumas, including neural ODEs, universal differential equations, mixed-effect neural networks, and deep non-linear mixed-effect models.
4. Engage in a practical case study to apply skills in DeepPumas modeling and gain insights into real-world data analysis.

Target Audience:
This workshop is ideal for pharmacometricians, data scientists, and researchers involved in drug development who want to enhance their skills in NLME modeling and learn about the integration of machine learning into pharmacometrics. Prior experience with modeling tools and a basic understanding of pharmacometrics are recommended.

Pricing: $1800 for Industry, $500 for Academia

Bring Your Team, Save More! Groups of 3+ qualify for a group discount. Email [email protected] to inquire.

Agenda:

Day 1: Utilizing Pumas for Non-linear Mixed-Effect Modeling

• Welcome and Introduction to the Workshop
• Overview of Non-linear Mixed-Effect Models: A Refresher on Key Concepts and Recent Developments
• Introduction to Pumas: A Comprehensive Tool for NLME Modeling
• Practical Session: Getting Started with Pumas for NLME Modeling
• Hands-On Exercises: Building Basic NLME Models in Pumas
• Diagnostics, analysis, plotting, and reporting.
• Q&A and Discussion: Addressing Specific Modeling Challenges and Use Cases

Day 2: Machine Learning in Drug Development and Pharmacokinetics

• Introduction to Machine Learning in Drug Development
• Basic supervised machine learning and neural networks
• Hands-on
• Bridging the language gap between machine learning and NLME modelling.
• Review of Key Papers and Studies in ML for Pharmacokinetics and Pharmacodynamics
• Case Studies of Machine Learning Applications in Pharmacokinetics
• Practical Session: Implementing Basic Machine Learning Models in Pharmacokinetics
• Group Discussions: Challenges and Opportunities

Day 3: DeepPumas – Dynamical Systems and Mixed-Effect Neural Networks

• Introduction to Neural Ordinary Differential Equations (Neural ODEs)
• Universal Differential Equations: Concepts and Applications
• Practical Session: Implementing Neural ODEs and Universal Differential Equations in DeepPumas
• Mixed-Effect Neural Networks: using random effects in supervised machine learning.
• Practical session: Using mixed-effect neural networks in DeepPumas
• Interactive Q&A and Case Study Discussions

Day 4: Deep Non-linear Mixed-Effect Models

• Introduction to Deep Non-linear Mixed-Effect Models (DeepNLME)
• Concepts and Applications of Deep NLME Models
• Practical tips-and-tricks.
• Practical Session: Building and Testing Deep Non-linear Mixed-Effect Models in Pumas
• Areas of application: Cool tool, now what?
• Model discovery
• Complex covariates
• Longitudinal biomakers
• Open Discussion

Day 5: Case Study and Independent Exercise

• Here's a data set - have at it!
• Introduction to Case Study Exercise: Dataset and Problem Statement
• Hands-On: Participants Conduct DeepPumas/Deep NLME Modeling
• Final Presentations: Sharing Findings and Insights
• Closing Remarks and Feedback

Host or Publisher Pumas-AI, Inc.