Emerging evidence suggests that metabolic products formed by gut bacteria directly impact on human health and disease. In this study, the authors proposed that the microbiome community is more predictive of host dysbiosis than the community structures themselves, particularly for cross-donor analysis. Also, different statistical and machine learning strategies were applied to characterize four types of microbiome features, i.e., taxonomy, community enzyme function profiles, and total and secondary community metabolome, and to compare their capabilities in predicting dysbiosis. Major Compulsory Revisions: 1. It is difficult to understand how the community enzyme function profile and metabolome are modeled from taxonomic relative abundance. Brief introduction or illustration of the process including methods and tools are necessary for readers to understand the 4 types of big data at the beginning.

1. As indicated by author, the limitation of this work was to use datasets from only two donors. For validation, it is necessary to dig further about their findings of metabolome through literature, and support their points in the discussion “Pathways for vitamin biosynthesis, protein and fatty acid digestion, and potential virulence factors were found to be significantly enriched for the predictive microbiome features.”

Minor Essential Revisions: 1. Bray-Curtis dissimilarity indices Section: Line 202: please provide the cited reference to the reference list. 2. Figure legend 1: Line 687: please provide the cited reference to the reference list.

Level of interest An article of importance in its field Quality of written English Acceptable Statistical review No, the manuscript does not need to be seen by a statistician. Declaration of competing interests I declare that I have no competing interests.

Authors' response to reviews: (http://www.gigasciencejournal.com/imedia/9870863771812193_comment.pdf)

The authors' revisions have addressed all my questions and concerns. I have no additional concerns. Level of interest An article of importance in its field Quality of written English Acceptable Statistical review No, the manuscript does not need to be seen by a statistician. Declaration of competing interests I declare that I have no competing interests.