Modeling and mapping microbial diversity and function with marker genes, genomes and metagenomes

As microbiomics becomes ever more metagenomics-based, principled methods of binning data for maximum diagnostic and predictive value will be ever more useful. Current species definitions remain problematic, and the extent of genomic and phenomic clustering and the processes that might be responsible for cluster formation are very poorly known. We use existing data and privileged early access to MLST, genomic and metagenomic data from collaborators to ask: (1) When should we consider setting aside "species" and what does this mean for biodiversity research, especially as it relates to health issues? (2) How does one study microbial biodiversity free from the preconception that genes will necessarily exist only within discrete species-like structures? (3) What new computational tools and methods will be required?