PopED lite an easy to use experimental design software for preclinical studies
Yasunori Aoki (1), Monika Sundqvist (2), Carl Whatling (3), Anna Rönnborg (3), Peter Gennemark (2), and Andrew C. Hooker (1)
(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (2) CVGI iMED DMPK AstraZeneca R&D, Sweden, (3) CVGI iMED Bioscience AstraZeneca R&D, Sweden
Objectives: In drug discovery, accurate and precise estimation of drug potency from in vivo data is crucial to select the right compound. Proper design of experiments is a fundamental requirement to obtain such estimates. Through this work, we wish to investigate how optimization of the experimental design can be utilized in compound selection studies, and then generalize our findings with a software tool dedicated for drug discovery applications to automate and accelerate the experimental design optimization processes.
Methods: In order to understand the challenges in designing the in vivo studies for drug discovery applications, we have chosen an ongoing compound selection study at AstraZeneca as a case study and used optimization techniques when designing new experiments.
For each new experiment, the experimental design was optimized using the experimental design optimization software PopED [1] and the theoretical optimal experimental design was suggested to the project team. The design was modified considering all the practical aspects of the experiment, and the changes were re-evaluated using PopED. Then, the experiment was conducted and once the experimental results were available, the experimental design was again evaluated. In addition, possible improvements of the design optimization processes for new experiments were explored.
Results: The study shows that the optimization of both sampling time and dosing scheme, especially the latter, can increase the accuracy of the estimation of the potency of the drug as well as other drug related parameters. Also, it is important to identify the practical experimental design constraints, such as the possible timeframe of observations, the possible time interval between observations, or the total amount of the compound available, that can influence the optimal experimental design, and incorporate these constraints into the design optimization algorithm.
Conclusions: In order to automate and accelerate the experimental design processes, we have implemented a simplified version of PopED that can be used in preclinical studies. This software features rich graphical representation of both the model simulations and the accuracy of the model parameter estimates together with an easy to use graphical user interface.
References:
[1] Nyberg, J., et al. PopED: An extended, parallelized, nonlinear mixed effects models optimal design tool. Computer Methods and Programs in Biomedicine, 2012. In Press.
