Hybrid Genetic Algorithm Approaches to Model Selection

A newly implemented version of an established Genetic Algorithm based approach using NONMEM with some modifications into an R-shiny package that provides the same functionality with additional flexibility is presented.
Some refinements have been made in the selection process for evaluating models compared with the original evolutionary algorithm. Several population analyses using the updated algorithm are compared with results obtained using classical stepwise approaches in xenograft tumor growth profiles and population pharmacokinetic analyses