Overall the work represents a valuable addition to the open-source software at the disposal of bioinformaticians in the field of microbial ecology. It falls short of opening the doors to biologists only by the fact that the successful installation of the software is dependent on so many other (open-source) packages, including QIIME (which in turn has multiple dependencies). Unfortunately I was unable to get the installation working on my machine due to these dependencies so I have been unable to test the command line software. That said, I did test the interactive user interface with the test data as supplied. It is indeed a powerful visualisation tool and allows the user to dissect and display the results in many ways, hence my recommendation to publish the article with the minor revisions listed below:

1. Figures: At the end of the main section (just before the conclusions) there is mention of Fig 1B-C and Fig 1C-D, These figures either dont exist, or they have been renamed as (Fig 1 B-1 and Fig 1 B-2) since the section was written?

2. The legend of the Figure needs to match the figure,
   e.g.:

2a.- "(A) Raw results of principal coordinates analysis (PCoA), very difficult to interpret, shown alongside EMPeror’s GUI that allows the user to alter visual aspects of the dataset." There is no picture of the GUI in the figure?

2b - "(B) The same data colored by body habitat; (C) the same the data ..." There is no "B" or "C" in the figure, its B-1 and B-2 (or there is some of the image missing?)

1. in the background section, the first mention of "Fig. A" should be changed to "Fig. 1A" or whatever is the correct figure after the above changes have been made.

2. Citations 9 and 10 not used in text, either remove from ref list or add into text.

In addition to the above minor revisions there a couple of discretionary revisions that the authors may consider:

1. Near the start of the section titled "EMPEROR" it might be worth adding a paragraph on how to get and install the software (mentioning its dependencies on QIIME and python modules/packages etc).

2. If space allows, it would be nice to have a brief summary of the 4 datasets included in the combined summary example with a comment on the potential dangers of integrating different datasets. I realise that could be an entire article on its own, but I think its worth pointing out to people that just because this kind of meta-analysis is POSSIBLE doesn't mean it should be done to everything, and only datasets with appropriate metadata should be compared.

typo:

1. explaind -> explained

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 am a current employee of GigaDB, which is a subsidiary of GigaScience Journal.

The revisions are a great improvement, especially the addition of the table, there is one thing I noted about the table, the total number of samples listed in the table is 5629, the number of samples quoted in the text is 5740, I think this discrepancy needs some attention. either an explanation as to why the difference, or checking the numbers are correct.

The addition of the files to GigaDB should allow even more users to access the tools and see how the examples were generated, the DOI reference will need to be added to the text and citations in the appropriate place.

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 am a current employee of GigaDB, which is a subsidiary of GigaScience Journal.

Chris