Improved sampling of diverse environments (e.g. ocean, soil, sediment, and a range of hosts) combined with advances in the development and application of ultra-high throughput sequence methodologies are set to accelerate the pace at which new metagenomes are generated. The M3 & Biosharing SIG explored the latest concepts, algorithms, tools, informatic pipelines, databases and standards that are being developed to cope with the analysis of vast quantities of metagenomic data. It also sought to facilitate a broader dialogue among funders, journals, standards developers, technology developers and researchers on the critical issue of data sharing within the metagenomics community and beyond through the inaugural meeting of the BioSharing community (See http://biosharing.org for a list of participating communities). Through two days of invited and contributed talks, panel discussions, we aimed to highlight scientific advances in these fields and identify core computational challenges facing the wider community.

Because the pace of genomic and metagenomic sequencing projects [¹] is rapidly increasing, and will only accelerate as the application of ultra-high-throughput methods becomes more wide spread, the role of standards is becoming ever more important to scientific progress and data sharing. The Genomic Standards Consortium (GSC) is an international working body with the mission of working towards richer descriptions of our collection of genomes and metagenomes through the development of standards and tools for supporting compliance and exchange of contextual information [²].

This report summarizes the proceedings of the Metagenomics, Metadata and MetaAnalysis (M3) Special Interest Group at ISMB 2010. Special Interest Group (SIG) meetings at the ISMB are a special way to bring together computational researchers interested in a particular topic. In establishing this new SIG, the GSC hopes to engage a wider range of bioinformatics researchers in thinking about standards. This SIG was named M3 to cover the important intersections between the ongoing explosion of data (Metagenomics) and the ever growing need to support richer stores of associated contextual data (Metadata) to improve our ability to interpret and comparing findings across large collections of independent studies (MetaAnalysis).

The agenda of the M3 SIG was designed to focus on the intersections of science and standards and policy. The first day of the SIG was dedicated to metagenomic science. The first session contained five plenary talks by speakers championing each of an expanded set of five Ms or M5 (Metagenomics, Metadata, MetaAnalysis, Models and MetaInfrastructure).

Ed DeLong (Massachusetts Institute of Technology) opened the session by talking about the promise of metagenomics for shedding light on community processes with his talk entitled Understanding Microbes through Metagenomics. This was followed by a talk from Michael Ashburner (University of Cambridge), a founder of the Gene Ontology and co-coordinator of the OBO Foundry on the power of ontologies. Phil Hugenholtz (Joint Genomes Institute) gave an overview of the Microbial Earth project, an ambitious effort to sequence over 9,000 type strains to fill in the tree of life.

Eric Alm (Massachusetts Institute of Technology) and Folker Meyer (Argonne National Laboratory), the final two plenary speakers focused on models and metainfrastructure. The vision of M5 is to bring democratized access to computation. The theme of how to bring the vision of an M5 platform to life was the theme of the final Panel session. Before lunch, Dietlind Gerloff (UC Santa Cruz) made a call for new collaborators in her flash talk Sharing ‘raw’ experimental data between biological laboratories in a new agent-based infrastructure OpenKnowledge.

After lunch, Session II continued with seven selected talks (from contributed abstracts) followed by a talk about the GSC to set the stage for the Panel session. The first pair of talks described software platforms. Shulei Sun (UC San Diego) talked about the latest release of the CAMERA 2.0 Workflow System. Daniel H. Huson (Tübingen University) then talked about functional analysis and comparison of meta-genomes, -transcriptomes and -proteomes using MEGAN4.

The next pair of talks described new approaches to understanding the content of metagenomic datasets. In the first talk, Alexandre Lomsadze (Georgia Institute of Technology) described an improved method for finding genes in metagenomic sequences. Gail Rosen (Drexel University) then talked about her new method of detecting Novel Species and Genera from Short Reads.

The focus then shifted to analysis of novel metagenomic datasets. The first talk described the generation of a dataset, the mobile gene pool metagenome from waste water, and was presented by Antonia P. Mayer (University of Lausanne).Riccardo Percudani (University of Parma) then described an analysis of human gut microbiome data that reveals an alternative pathway for urate degradation - a pathway for breaking down uric acid lost from the lineage leading to humans but potentially found in human gut bacteria.

The final pair of speakers brought the topic back around to the need for standards and efforts to build consensus. Dirk Gevers (Broad Institute) of the Human Microbiome Project then talked about the need to standardize protocols across laboratories. Renzo Kottmann (Max Planck Institute Bremen), a member of the GSC Board, then closed out the session by describing the work of the GSC in a talk entitled The Genomic Standards Consortium: A Community of Individuals.

After the coffee break, Folker Meyer(Argonne National Laboratory) and Sarah Hunter(European Bioinformatics Institute) chaired an M5 panel discussion involving the plenary speakers from the first day. This discussion advanced a group vision of a collaborative, computational infrastructure to advance research in this area. More information on the GSC current strategy of the GSC about M5 can be found on the M5 homepage [³].

In summary, the second M3 SIG meeting at ISMB brought together a wide range of researchers to share results and discuss the need for community-based solutions in the field of metagenomics. The GSC will continue to build on the M3 workshop series, with the next workshop held at July 2010 at biannual meeting of the International Society for Microbial Ecology. As a result of this workshop, the GSC helped to engage further with the international bioinformatics community, articulate more clearly the goals of the M5 roadmap and ultimately launch the BioSharing community [⁴]. The formation of the BioSharing community, including the GSC as a leading community, will undoubtedly aid the GSC in fulfilling its mission of improving data sharing and re-use in the specific domain of genomics and metagenomics.

Many thanks to all invited and selected speakers and everyone that participated in this SIG meeting. We gratefully acknowledge the support from the US National Science Foundation (NSF) grant RCN4GSC, DBI-0840989. The work of Lynette Hirschman was funded in part by NSF grant IIS-0844419.