On September 14, Dr. Nathan Teuscher presented an informative webinar focused on new features and enhancements in Phoenix® 7.0, which was released on August 10. This blog post summarizes the topics discussed during the presentation.
Nathan began his presentation with an overview of new features in Phoenix 7.0 including:
- A new graphics engine for high resolution plots and figures
- Enhancement of the WinNonlin® engine using parallel computing for non-compartmental analysis (NCA) which halves analysis time
- Implementation of a new Data Layer which improves overall performance, particularly for large data sets
- New integrated grid and cloud support for nonlinear mixed effects modeling (NLME™) jobs
- New features added to NLME make it the most user-friendly software for population PK/PD studies
New graphics engine
Phoenix 7.0’s new graphics engine will benefit WinNonlin, NLME and IVIVC Toolkit users. As Nathan demonstrated, users are now able to generate high resolution plots and figures for publications, poster presentations, and regulatory reports, eliminating the need for third-party graphing tools. Phoenix 7.0 plots are crisp, clear and have high native resolutions up to 1200 dpi; standard image resolution for publications are 300 to 600 dpi. In addition, users can now control the location of legends, including placement inside the plot area. Adding to the enhancements for out-of-the-box publication-quality graphics, is shading of the visual predictive check (VPC) plot produced in NLME. The shaded areas represent prediction intervals and are used to quickly identify observed data that fall within the prediction interval. This plot is produced automatically when a VPC is run in NLME. We have also added the ability to create offset plots which compare two items by displaying both the individual data and mean data with its standard deviation. Before having this feature in Phoenix 7.0, these complex plots were only available with third party graphics tools. But Phoenix 7.0 can produce these easily and quickly using the comfortable plotting interface.
The industry-standard WinNonlin engine has been redesigned using parallel computing technology to conduct NCA twice as fast as our previous version. This improvement will benefit WinNonlin users that perform NCA for large animal studies, toxicology studies, or bioequivalence trials. Phoenix 7.0 takes advantage of multiple cores on local computers to run all subjects in parallel. In the example given using 800 PK profiles, Nathan demonstrated how NCA analysis time is reduced by 50% as compared to Phoenix 1.4 where each PK profile is evaluated sequentially. This feature saves time, particularly for large NCA projects with many subjects and analytes.
NCA set up and execution remains the same. Phoenix automatically detects multiple processors and then parallelizes the analysis to halve your analysis time!
New data layer
Phoenix 7.0’s new data layer reduces memory utilization by 90% compared to the previous version. It uses memory and hard drive to provide a much more stable application, especially for large datasets. This update improves the user experience.
Grid and cloud computing with NLME
Phoenix 7.0 provides first-in-class integrated support for execution of NLME jobs on remote/parallel compute platforms including MPI clusters, Linux computers, and Linux grids using Torque or SGE. Users can leverage an unlimited number of cores. Grids hosted on Amazon web services are also supported. This is an enormous advance for NLME users as they can offload long runs to cloud computing resources.
Phoenix NLME is the first PK/PD modeling application that supports grid computing out of the box – check out the Frequently Asked Questions for Grid and Cloud Computing to learn everything you wanted to know about Grid computing with NLME.
Additional new features and enhancements in NLME
Phoenix 7.0 implements new features for NLME that enhance the user’s experience and make it the Modern PK/PD Modeling Software for Today’s Scientists. New diagnostic messaging pinpoints model coding and setup errors and provides solutions to get you back-on-track faster. Informative error messages benefit all NLME modelers, and new modelers in particular. Other new features include:
- Implementation of delay differential equations (DDEs) with a single command eliminating the need for multiple lines of Fortran code. DDEs are used to model delays between drug administration and the onset of biological activity.
- Categorized model variables can be input using text values such as Male and Female rather than having to convert them to numerical values first. This feature reduces time spent formatting datasets before analysis.
- Updated drug infusion setting to model either the rate of the infusion or the duration of the infusion.
- Improved handling of bounds and standard error calculations in NLME.
- Individual Post-hoc parameters have been added as standard output for each Phoenix Model.
These new features and enhancements were inspired by user feedback. They make the Phoenix workbench the most efficient tool available for pharmacokineticists and pharmacometricians.
If you weren’t able to attend the webinar, no worries. You can still view it in the archived webinars section of our Resource Center or subscribe to our YouTube channel. Have a look and find out all that is new in Phoenix 7.0