Full Agenda - ICON plc
Full Agenda - ICON plc
Full Agenda - ICON plc
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Introductory NONMEM & PDx-Pop & New and Advanced Features of NONMEM 7 and<br />
PDx-POP 5 Workshop<br />
Date of Meeting: 9-11 April 2013 & 10-12 September 2013<br />
Location:<br />
Presenters:<br />
Ellicott City, Maryland<br />
William Bachman, PhD; Robert Bauer, PhD<br />
<strong>Agenda</strong> Items:<br />
DAY 1 - "Introductory NONMEM & PDx-Pop" - Presentation: Instructor Time<br />
1. Course Introduction William Bachman, PhD<br />
9:00 –<br />
9:15 AM<br />
2.<br />
Introduction to Population Pharmacokinetics (2hrs)<br />
- Classical Population Analysis Principles/Methodologies<br />
- Sums of Squares Objective Functions<br />
- Extended Least Squares<br />
- Fixed and Random effects<br />
- Population Analysis Methods<br />
- First Order<br />
- First Order Conditional Estimation<br />
- Model Development Validation<br />
- data exploration<br />
- structural and error model development<br />
- goodness of fit diagnostics<br />
- diagnostic plots<br />
Robert Bauer, PhD<br />
William Bachman, PhD<br />
9:15 –<br />
11:15 AM<br />
3. Coffee Break<br />
11:15 –<br />
11:30 AM<br />
4.<br />
Analysis of Data from One Subject (3hrs)<br />
- Example (PK, Variance Models)<br />
- Introduction to NONMEM Files and Analysis<br />
Robert Bauer, PhD 11:30 –<br />
12:30 AM<br />
5. Lunch<br />
6. Analysis of Data from One Subject (continued)<br />
7. Break<br />
12:30 –<br />
1:30 AM<br />
Robert Bauer, PhD 1:30 –<br />
3:30 PM<br />
3:30 –<br />
3:45 PM<br />
8. Population Analysis with No Covariates (1.75hrs) William Bachman, PhD<br />
3:45 –<br />
5:30 PM<br />
2012 Page 1 of 4 Prepared by: L. Wilhelm
2012<br />
DAY 2 – "Introductory NONMEM & PDx-Pop" - Presentation: Presenter Time<br />
1. Population Analysis with No Covariates (continued) William Bachman, PhD<br />
8:30 –<br />
9:00 AM<br />
2.<br />
Population Analysis with Covariates (3.25hrs)<br />
- NONMEM Files<br />
- Exploratory Analysis<br />
- Model Building Example<br />
Robert Bauer, PhD 9:00 –<br />
10:00 AM<br />
3. Break<br />
4. Population Analysis with Covariates (continued)<br />
5. Lunch<br />
6. Population Analysis with Covariates (continued)<br />
10:00 –<br />
10:15 AM<br />
Robert Bauer, PhD 10:15 –<br />
11:30 AM<br />
11:30 –<br />
12:30 AM<br />
Robert Bauer, PhD 12:30 –<br />
1:30 PM<br />
7.<br />
Introduction to Model Evaluation Methods<br />
- User-Written Models in NONMEM (1.75hrs)<br />
* A PK-PD Example<br />
* NONMEM Files<br />
William Bachman, PhD<br />
1:30 –<br />
2:30 PM<br />
8. Break<br />
2:30 –<br />
2:45 PM<br />
9.<br />
User-Written Models (continued)<br />
PK-PD Model Problem<br />
Robert Bauer, PhD 2:45 –<br />
3:30 PM<br />
10. Population Analysis Final Example & Practice Session William Bachman, PhD<br />
3:30 –<br />
5:00 PM<br />
DAY 3 - "New and Advanced Features of NONMEM 7 and<br />
PDx-POP 5" – Presentation:<br />
1. Introduction Robert Bauer, PhD<br />
2.<br />
Modifications and Enhancements to NONMEM 7<br />
- Conversion of Fortran 77 to Fortran 90/95<br />
- Centralized Error Processing<br />
- Improvements in Gradient Methods<br />
- Decreased Incidence of Estimation Failure Due to Numerical<br />
Problems<br />
- Added Option to Specify Step-Size for Gradient Calculation<br />
- Increased Number of Data Items and Label Lengths<br />
- Flexible Numerical Formats for Input and Output<br />
- Added Information in Standard Results File<br />
- Identifier Tags for Certain Sections<br />
- Shrinkage of Variance<br />
- Additional Output Files Easily Readable by Post-Processing Software<br />
William Bachman, PhD<br />
8:30 –<br />
8:45 AM<br />
8:45 –<br />
9:30 AM<br />
2012 Page 2 of 4 Prepared By: L. Wilhelm
2012<br />
- Additional Weighted Residuals outputs (conditional, exact versions)<br />
3.<br />
New Methods Available in NONMEM 7, Theory, Overview<br />
- Monte Carlo Importance Sampling Expectation Maximization (EM)<br />
(IMP)<br />
- Markov Chain Monte Carlo (MCMC) Stochastic Approximation EM<br />
(SAEM)<br />
- Iterative Two Stage (ITS)<br />
Robert Bauer, PhD<br />
9:30 –<br />
10:00 AM<br />
4. Break<br />
10:00 –<br />
10:15 AM<br />
5.<br />
Mu Modeling<br />
- Model Modifications That Improve Efficiency of EM Methods (Mu<br />
Modeling)<br />
Robert Bauer, PhD<br />
10:15 –<br />
11:00 AM<br />
6.<br />
Examples for EM Methods (hands-on)<br />
- Basic two compartment model problem, incorporate Mu Model<br />
Robert Bauer, PhD<br />
11:00 –<br />
11:45 AM<br />
7. Lunch<br />
11:45 –<br />
12:45 PM<br />
8.<br />
Examples for EM Methods (hands-on)<br />
- Two compartment model with age and gender covariates Robert Bauer, PhD<br />
12:45 –<br />
1:45 PM<br />
9.<br />
Bayesian Analysis (hands-on)<br />
- MCMC Bayesian Analysis (BAYES)<br />
- Prior information for MCMC Bayesian Analysis<br />
- Revisit two compartment model, adding Bayesian analysis<br />
Robert Bauer, PhD<br />
1:45 –<br />
2:45 PM<br />
10. Break<br />
11.<br />
More Examples with EM and Bayesian Analysis<br />
- Population mixture model problem<br />
- Interoccasion variability problem<br />
- Categorical data problem<br />
2:45 –<br />
3:00 PM<br />
3:00-3:30<br />
12.<br />
13.<br />
Additional Considerations for EM and Bayesian Analysis<br />
- Termination Testing<br />
- Making Numerically Stable Models<br />
- Gibbs vs. Metropolis-Hastings<br />
Creating Random Initial Parameters for Multiple Chains (CHAIN)<br />
- Chain command syntax<br />
- Thetas: univariate or normal randomization<br />
- Sigmas: univariate randomization<br />
- Omegas: Wishart randomization<br />
- Using random samples for immediate, or later problems<br />
- Creating random initial values example<br />
Robert Bauer, PhD<br />
3:30-4:00<br />
4:00 –<br />
4:30 PM<br />
14.<br />
PDx-Pop Interface For NONMEM 7<br />
- Real-Time Graphical Monitoring of Objective Function<br />
- Interaction with NONMEM Run<br />
- Toggle Switch for Console Printing of Iterations<br />
- Switch to End a Problem Gracefully<br />
- Switch to End a NONMEM Gracefully<br />
William Bachman, PhD<br />
4:30 –<br />
5:00 PM<br />
2012 Page 3 of 4 Prepared By: L. Wilhelm
2012<br />
- Extended Summary Output<br />
- Graphical Display of Parameter Sampling History (BAYES)<br />
- Setting Up and Running Multiple Analysis Chains Simultaneously<br />
- Graphical and Tabular Summary of Multiple Analysis Chains (BAYES)<br />
- nitial Parameters Variation Test<br />
- PDx-Pop on Linux and MAC OS X<br />
15. Question and Answer Session, Demonstrations 5:00 PM<br />
2012 Page 4 of 4 Prepared By: L. Wilhelm