Hi, Does Phonex Winnolin still have scripting language feature? I used to run SAS concurrently with Winnonlin 5.x via scripting language (the psc files) to do the following (in steps): 1) Use SAS to import in the BA/BE data (subject, period, Treatment, seq, time, concentration) 2) Access Winnonlin 5.x within SAS by Launching WinNolin and execute a script 3) According to the script, Winnolin proceeds to analyze the dataset, compute the PK parameters and export the results in excel. 4) Once the script finishes, SAS would resume, import the winonlin data (excel format), and perform the stat analysis (GLM) and compute 90% CIs. Is the above possible with Phoenix? Thanks John
Hi John, Phoenix does have an scripting language but it is completely different than the WNL5 scripting language. Phoenix can be scripted via the Phoenix API. You can contact Pharsight Sales if you are interested in getting additional information. However, you can also do a similar workflow as you mentioned via a SAS object in a Phoenix workflow too. To use that feature you would need Phoenix connect. It would work as follows: 1. Create workflow 2. Import input data 3. Run NCA via PHX WNL 4. Run a SAS object to the GLM analysis 5. Tabulate data in PHX (for example) You can create a template workflow and execute this with new datasets. This would not be ‘scripted’ but very simple to use. I am attaching a project that has a SAS object for you to take a look. You won’t be able to run it without a PHX Connect license but you can see the results. Ana [file name=NCA_BE_and_SAS.phxproj size=1387716]/extranet/media/kunena/attachments/legacy/files/NCA_BE_and_SAS.phxproj[/file]NCA_BE_and_SAS.phxproj (1.32 MB)
Hi Ana, replying to the last of the manifold of posts the board’s scripts managed to produce. 
Hey, what is this magick API? We are still exploring methods to detect outliers in replicate design studies (which is EMA’s requirement). Internally studentized residuals – no problem (Residual / Standard Error). Probably a better approach would be by externally studentized residuals Studentized residual - Wikipedia] here Essentially that would mean rerun the BE-model excluding one subject at a time. Nightmare. I even considered giving it a try in WNL5.3, but the setup of the model there is nasty.