Pirana Modeling Workbench
Organizing and managing pharmacometric models can be an inefficient and time-consuming process. Pirana’s workbench tools provide modelers with structure to facilitate the iterative processes used to create population PK/PD models and perform simulations resulting in better organization and more efficient analysis of population PK/PD results.
Reduce Model Development Time by 20-30% with New Machine Learning Based Automated Model Selection
The new Darwin Model Search feature in Pirana provides a streamlined user interface to pyDarwin, a cutting edge open-source Python package funded by the US FDA that implements a number of machine learning algorithms for automated pharmacometrics (PMX) model selection. The best model selected by a Darwin search is fully compatible with the existing suite of Pirana tools for PMX workflow management. Learn more about pyDarwin here.
The Pirana Modeling Workbench includes:
- Darwin automated model search feature
- Model templates and wizards
- Library of goodness-of-fit plots (R)
- Model translation tools
- Seamless access to open source tools to interface with pop PK/PD modeling tools for visual predictive check, covariate modeling, and bootstraps
Pirana works with R speaks NLME, NONMEM, R, and Perl Speaks NONMEM (PSN). Pirana is supported on Windows and Linux operating systems.
Organize and manage models
Analogous to a lab notebook, Pirana creates the necessary infrastructure to manage, create, track, and perform PMX data analyses.
It delivers faster, more efficient modeling and simulation by connecting multiple tools for optimizing a pharmacometrics workflow.
Pirana provides an automated folder structure for the storage of analytical results, adding order and traceability to complex analyses.
Analyze data and generate reports
Pirana offers a quick and easy graphical interface to R plotting tools and supports the generation of plots in R using PSN and Xpose or custom user scripts. Our powerful reporting tools provide options to run turnkey reports in various formats, including Word, html, plain text, LaTeX, and Excel. Model output can also be processed using the ‘Model Results’ Shiny application which allows point and click creation and customization of graphical and tabular outputs which can be rendered as a R markdown document in pdf, word, or html format. The underlying R code is returned to the user as a “how-to” example, or for further customization to produce the exact output demanded by the problem at hand.
Visualize the model decision tree
The Visual Run Record in Pirana documents the model development process through a unique interactive visualization.
A summary of the model selection thought process displays how models were evaluated, and how and why the final model was chosen.
Pirana also provides fast and easy model evaluation and side-by-side model diagnostics to aid in the model development process.
