I was finally going to cast my vote on the meta-pipeline project, @Evangelos Karatzas. Unfortunately, I see drawbacks in all that are a bit difficult to circumvent: too old, no metaT and more. At the Linnaeus University we actually have a dataset that would work with the not insignificant drawback of not being clinical nor an experiment. At a station in the Baltic Sea we have since 2011 collected 16S amplicons (~every two weeks; three filter setups) and during two periods (ca. two years each), also metaG and metaT. No metaP, at least not published yet. For the metaT and metaG there are some inconsistencies regarding filtering setup and I'm a little unsure of the publication status of some data but the majority is publicly available or will very soon be. Personally, I of course think this is a fantastic dataset, but impact of a paper would likely be greater with some medical angle, preferably patients and controls.