Thus, targeting, e.g., one species per genus might not be a wise choice, even if all genera were monophyletic. Only for the species rank, microbial taxonomy has firmly established a criterion related to character divergence, namely the DNA-DNA hybridization (DDH), traditionally selleck inhibitor conducted in the wet lab [6] but more recently using genome-sequence based, digital replacements [47]. DDH, however, is a similarity method, whereas more similar organisms are not necessarily more closely related [3,28,48,53].A further problem with the approach to generate one genome per taxon (of a chosen taxonomic rank) is that the number of genomes to be sequenced would not depend on the available project resources but on the number of taxa. Neither a ranking within nor between those taxa would be provided.

The same difficulty would arise if non-hierarchical sequence clustering was used, followed by selecting one organism per cluster, even though here the number of clusters could be chosen (using, e.g., K-means partitioning [54]) and thus adapted to the project’s needs. But in contrast to the suggested phylogeny-based scoring, no continuous ranking would be provided, and re-clustering would be necessary after each change in the number of target genomes. Using trees with branch lengths for target selection thus seems to be the best choice, and the ease with which scoring systems such as the one described here can be inferred from phylogenies renders such methods rather promising. Acknowledgements Cordial thanks are addressed to Dorothea Gleim and the DSMZ curators for providing information about all GEBA target strains, to A.

Fiebig for information on the Roseobacter clade, and to J. Meier-Kolthoff (all at DSMZ), for other helpful comments. The work on the Roseobacter clade has been funded by the German Research Foundation (DFG) SFB/TRR 51, which is gratefully acknowledged.
An enthusiastic group of individuals with interests in functional metagenomics research convened in St. Jacobs, Ontario, Canada, for two days of intensive discussions on metagenomics. The participants included 26 attendees from academia, funding agencies and industry. The 1st International Functional Metagenomics Workshop (IFMW) was organized and hosted by researchers from the University of Waterloo and funded by the University of Waterloo, Ontario Genomics Institute (OGI), and Natural Sciences and Engineering Research Council of Canada (NSERC).

Its purpose was to identify challenges and opportunities and determine the best ways for the community to work together to achieve its goals. An important aspect of these discussions was the implementation of mechanisms for sharing resources such as metagenomic libraries, Carfilzomib and the establishment of standards for library construction and data collection. A number of natural relationships have already formed among practitioners of functional metagenomics.