Phoenix NLME: Nonlinear Mixed Effects Modeling

Phoenix® NLME is a powerful population modeling and simulation software for experienced as well as novice PK/PD scientists. With an intuitive user interface, it is an integrated solution that allows researchers to pre-process, analyze and post-process modeling projects all within one system and without having to write code. With it, researchers can:

  • Perform exploratory analysis
  • Develop, evaluate and compare many models
  • Post-process results and create figures for reporting


  • Saves time spent learning complex coding: Phoenix NLME’s intuitive graphical user interface shortens the learning curve for drug development scientists who want to use population modeling methods. It offers the flexibility to construct models using models from a library, graphically, or by manually coding for more advanced users.
  • Leverages powerful modeling capabilities: Phoenix NLME contains a variety of estimation engines including the fast and highly accurate QRPEM.
  • Gains in efficiency: Organize analysis steps in a colorful, easy-to-understand workflow. Save time on commonly performed tasks by creating reusable templates from workflows.
  • Simplifies Quality Assurance: Users can send the project file, which contains all analysis steps and documentation, to another researcher, QA staff, or a regulatory agency for easy re-analysis.
  • Enables compliance: Phoenix NLME provides technical controls for 21 CFR Part 11 compliance when used in conjunction with the Phoenix Knowledgebase Server™.
  • Supports Validation Environments: A validation suite (sold separately) for Phoenix NLME supports onsite validation.


Modeling Capabilities

  • Wide set of optimization engines for nonlinear mixed effects modeling, including an EM (expectation maximization) method (QRPEM)
  • Provides comprehensive output of tables and graphs to evaluate and develop models
  • Includes tools for diagnostic model checks like visual predictive check and bootstrap options
  • Has automated algorithms to evaluate the need for covariates in models and a tool to compare models
  • Contains a unique start up procedure for initial estimates
  • QRPEM engine handles inter-occasion variability
  • Users can select many residual error models including two sigmas/epsilons

Computational Performance

  • Supports parallel execution on multiple processors within a single run
  • Employs dynamic memory allocation and therefore there are no static limits (e.g., number of observations per subject)
  • Uses the Argonne MPICH2, which allows optimizations to be run on multiple processors simultaneously
  • Takes advantage of modern dual- or quad-core processors, which results in significant speed gains
  • Contains the ability to stop a run before it is finished and provide results from the last completed iteration as part of the updates to Phoenix NLME 1.3
  • Supports dispatching jobs (including workflows) to execute remotely on a collection of Windows machines via the new Job Management System (JMS)

Contact us to learn why Phoenix NLME is the best in class solution for population PK/PD modeling and simulation!