Multivariate Data Analysis in Microbial Ecology - New Skin for the old Ceremony

The molecular biology revolution has a particularly strong impact in microbial ecology, as
molecular methods are now giving access to data that were previously impossible to obtain.
Soil microbial ecology studies are a good example of this situation. Knowledge of soil
bacterial diversity is of great interest, both from an applied agronomical perspective, and in
the framework of theoretical ecological models like the species-area relationship. Previously,
only culturable species could be studied, which represented an extremely low part of the total
bacterial community. Today, tools like DNA fingerprints, DNA microarrays, and
transcriptomic methods can be used directly on DNA extracts from bulk soil samples,
providing new insights into the diversity and functioning of bacterial soil communities.

However, while molecular tools have been rapidly apropriated by microbiologists, the
statistical methods needed to analyse the resulting huge amounts of numerical information
still represent an obstacle for many microbial ecologists. The R environment already plays an
important role in genomic data analysis thanks to the bioconductor project, but the statistical
methods needed for multivariate ecological data analysis are part of standard R packages, like
vegan and ade4. Although these packages are designed for plant or animal ecology, many of
their functions can be used to analyse molecular biology data sets. Furthermore, other
packages, like seqinr and made4 are very useful to bridge standard packages and genomic
data structures. Lastly, graphical user interfaces are also needed to help biologists master the
intricacies of some R functions.