The statistical challenges in analysing microbiome data are the following: they are compositional, contain a large number of zeroes, lthe ibrary size variability can be large, and they are strongly right-skewed. Another aspect is that microbiome data can vary greatly in all these characteristics. In addition, several different types of questions can be asked, necessitating in some cases the use of different methods. There is no lack of choice: alarge number ofdata processing methods are available and more methods and insights continue to be developed. Choosing the right methods to use on a particular data set thatmatch the research questions is no trivial task as no universally accepted approach has yet been defined.
WUR groups involved in microbiome research start from a biological angle and dont necessarily have the quantitative skills to make the best choices needed in analyzing these complex data, leading to a time consuming process of data analysis. The work from this project will result in a set of good practices and accompanying computing code. Thus, all WUR groups involved in microbiome research will profit from this project, greatly reducing the time they need to invest in data analysis.
|Effective start/end date||1/01/20 → 31/12/20|