Reproducibility in functional genomics and how to lie with statistics.
Misuse of statistics is considered one of the major reasons for lack of scientific reproducibility. Not only do the very same problems exist in many areas of bioinformatics, but they are also amplified by the highly dimensional nature of the data. I will give examples in which fundamental problems arise in bioinformatic data analysis and demonstrate an approach to test reproducibility in bioinformatics.
Host: Justyna Wolinska