A biologist, a statistician, and a bioinformatician walk into a conference room… and walk out with a great metagenomics project plan

Reviews of metagenomics analysis often emphasize the interdisciplinary and technical aspects of data analysis (Knief, 2014; Sharpton, 2014). How might these recommendations be implemented for future projects? In this opinion, I provide some areas to consider, especially in experimental design and full-data-cycle planning, and in expertise areas of value to metagenomics projects. This opinion is structured as a hypothetical conversation that reviews state-of-the art in these areas and brings out the various aspects of metagenomics project design. Let’s consider an example project in plant-microbe metagenomics—analysis of microbial metagenome functional genes that predict plant yield differences in a horticultural crop species. I’ve framed a discussion of key points as a conversation, perhaps at the third or fourth meeting, after participants have described their range of expertise.