Reproducible research

Monday, April 21th, 2008 by hinrich


Inspired by the work of Markus Rauschhaupt et al. (MCRestimate / Compendium) I have switched to using Sweave for my data analysis work. While there are certainly many other approaches, this concept seems to be quite useful as it combines the writing of R code for data analysis of various molecular profiling data with the writing of the report and the conclusions drawn from the analysis. What is especially nice is that one single piece of combined analysis code and documentation code generates together a report. At the time of execution of the file the R code is run, the anlysis is carried out and the report is written. All in all, one little step towards ensuring reproducible research in bioinformatics that will hopefully lead to better knowledge in molecular biology as every single step can be verified by external people.

Posted in Science