Peer Review Report For: Peccary, a metaprogramming open-source R platform to improve pharmacometrics efficiency [version 1; peer review: 1 approved, 1 approved with reservations]

Le résumé fourni par la source

Background: A pharmacometrics (PMx) workflow usually requires several software tools to cover all the steps from data analysis to model evaluation and simulations. However, these tools do not always communicate well together, compromising the efficiency of the whole process. Highly inspired by the markdown/pandoc system, we developed Peccary, an R package and its dedicated Shiny Application with the objective to accelerate the use of previously released R packages through various translation and metaprogramming processes. Methods: Peccary was developed with an agile method, progressively aggregating snippets of R code produced during real-life pharmacometrics works. Its first subpackage, PeccAnalysis, can be used to produce and evaluate various R code for population description (using table1 package), plot exploration (ggplot) and non-compartment analysis (pknca). The second subpackage, PeccaReverse, allows writing a structural model using a minimalist (simplified deSolve) syntax, before metaprogramming model simulations (using either deSolve or RxODE) and design evaluation (through PopED), along with performing various model syntax translations (e.g., into NONMEM, Monolix or nlmixr files). Finally, the third subpackage, PeccaResult, standardizes run outputs of several PMx software (NONMEM, Monolix, Adapt, nlmixr) to perform various diagnostic evaluation tasks. Results: The metaprogramming system used in PeccAnalysis and PeccaReverse has many advantages. First, it simplifies the use of previously mentioned packages (by reducing the required knowledge and the time needed to program the output creation). Second, it creates links between independent tools (for instance using the same inputs for several tasks). Third, the produced R code can be reviewed for eventual manual modification, verification (quality control) or traceable report integration. Conclusion: Overall, Peccary was successful in improving PMx efficiency by providing a Shiny R platform that can produce various outputs during live meetings while keeping the possibility to extract Peccary-independent R source code for further in-depth control.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.