Optimized design and analysis of preclinical intervention studies in vivo

Optimized design and analysis of preclinical intervention studies in vivo

We published our work on improving experiment design in preclinical in vivo experiments in Scientific Reports just now:

http://www.nature.com/articles/srep30723


I sure hope many experimenters will pick up at least the concept of how important sufficient sample sizes, good blinding, masking and randomization, as well as feasible post-intervention testing are. There's so many references out there telling how appalling the attrition rates in these experiments are at the moment, and I have personally also experienced this - preclinical experimenters should at least consult a biostatistician/-informatician on good design. Anyway, there's an R-package I maintain with this publication called 'hamlet' that's available on CRAN (comes additional machine learning trinkets):

https://cran.r-project.org/package=hamlet

-TDL, August 2016