The manuscript describes an alternative workflow for the processing of shotgun metagenomics and metatranscriptomic data, called ASaiM. ASaiM integrates multiple tools for the analysis and manipulation of raw metagenomics and metatranscriptomic data, that are available, both as single tools and combined in multiple pipelines, within the Galaxy workflow and with a Docker and conda support. ASaiM comes with a very impressive documentation and it is of high importance in the metagenomics community, where most of the analyses are carried out using in-house scripts that, as pointed out by the authors, hinder reproducibility.

However, several other metagenomics pipelines are already available: MG-RAST and the EBI metagenomics pipeline, that the authors briefly discuss in the Introduction, but also MOCAT2, MetAMOS, and another Galaxy metagenomic pipeline. How does ASaiM compare within this wider ecosystem? MOCAT2, for instance, comes with a set of preset parameters, stored in a single file, that already improve reproducibility, and the EBI metagenomics pipeline clearly shows the software version (e.g., https://www.ebi.ac.uk/metagenomics/pipelines/3.0), allowing provenance.

Also, the authors point out that the main problems in analysing metagenomics data are, first, the selection and configuration of the necessary tools, then the definition of the correct computational resources, and, finally, the definition of a correct analysis workflow. However, in this reviewer's opinion, ASaiM does not fully address these limitations. The authors implement about 25 tools for the processing of metagenomics data but give little explanation on the reasons these specific tools have been selected, or which tools should be used when multiple tools within the same class are available. Novices in the field would surely appreciate these pieces of information as a way to select the correct software for the problem at hand. Regarding the workflows included in ASaiM, one is a reimplementation of the EBI workflow, one cannot be used for analysing metagenomic shotgun data, and only one is novel (that this reviewer supposes is the one called very generally "ASaiM"). This reviewer would suggest the authors to focus more on describing this novel workflow, and to remove all the references to QIIME and Mothur tools (or to 16S data analysis in general) since these are not able to analyse shotgun metagenomics data and may generate confusion. For instance, it would be interesting to know how the workflow can be customised, whether default parameters are available and how they have been selected, and have more detailed and exhaustive information on time and computational requirements (and not only on two samples).

Also, it is not clear what improvements are brought by ASaiM and what are due to the usage of Galaxy (reproducibility, provenance, being user-friendly), or of HUMAnN2 (ability to infer the taxonomic profiles up to the species level, availability of genes and pathways abundances tables). For instance, how the proposed 'functional and taxonomic combination analysis' block differs with that proposed within the HUMAnN2 pipeline?

More in general, this reviewer's main concern regards the focus of the manuscript. Are the authors interested in presenting the Galaxy implementation of a variety of metagenomics tools? Or to present a novel reproducible pipeline for the analysis of metagenomics data? Are they interested in metagenomics or metagenetics (16S) analysis? In this reviewer's opinion, the manuscript would surely benefit in focusing on a single message, while additional features (such as the analysis of metagenetics data) should be only briefly mentioned.

The manuscript includes some imprecision, with several concepts repeated multiple times, and would surely benefit from a proofreading by a native speaker: 1. Lines 40-43. Metagenomics and metatranscriptomics techniques do not allow to get insight into metabolic components, but only on the inferred functions of the micro-organisms present in one sample (as done, for instance, by HUMAnN2). To measure the metabolic components, one should use another approach, namely metametabolomics. It is also not clear what 'phylogenetic properties' are. Do the authors mean taxonomical profiles? 2. Line 44. The authors mention 'high variability'. What is the feature showing this 'high variability'? 3. Line 52. Can the authors give examples of what they call 'computational resources specially for the metagenomics datasets'? 4. Line 140. What is a 'data reduction step'? 4. This reviewer suggests removing the 'Installation and running section' and simply refers to the documentation, as done in other cases.

Level of interest Please indicate how interesting you found the manuscript:
An article whose findings are important to those with closely related research interests

Quality of written English Please indicate the quality of language in the manuscript:
Needs some language corrections before being published

Declaration of competing interests Please complete a declaration of competing interests, considering the following questions: Have you in the past five years received reimbursements, fees, funding, or salary from an organisation that may in any way gain or lose financially from the publication of this manuscript, either now or in the future? Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this manuscript, either now or in the future? Do you hold or are you currently applying for any patents relating to the content of the manuscript? Have you received reimbursements, fees, funding, or salary from an organization that holds or has applied for patents relating to the content of the manuscript? Do you have any other financial competing interests? Do you have any non-financial competing interests in relation to this paper? If you can answer no to all of the above, write 'I declare that I have no competing interests' below. If your reply is yes to any, please give details below.
None.

I agree to the open peer review policy of the journal. I understand that my name will be included on my report to the authors and, if the manuscript is accepted for publication, my named report including any attachments I upload will be posted on the website along with the authors' responses. I agree for my report to be made available under an Open Access Creative Commons CC-BY license (http://creativecommons.org/licenses/by/4.0/). I understand that any comments which I do not wish to be included in my named report can be included as confidential comments to the editors, which will not be published.
I agree to the open peer review policy of the journal.

Authors' response to reviews: Reviewer #1

“Excellent paper, useful collections of tools, focused approach and well organized with great documentation.”

We thank the reviewer for this nice comment.

“Enough background for a software paper, my suggestion would be if you can mention a little more on metagenomics pipelines available on the main Galaxy server, in addition to an example of specialized Galaxy servers for metagenomics - for example the Metaphlan group they have such a specialized server: https://huttenhower.sph.harvard.edu/galaxy/”

We added in the introduction a sentence about the main Galaxy server and the metagenomics tools available there. We also mentioned the server of the Huttenhower lab. Moreover, we are in contact with the administrators of usegalaxy.org and we will ensure that all workflows and trainings will also work on their server.

“In addition I was really excited to see the provenance mentioned. Since the documentation is so extensive (and excellent!), perhaps the authors could add a section on how to save a docker container where data has been processed with their tool (also how to bundle the volumes with the data), so that the whole package and be distributed (and provide analysis provenance), to collaborators, with a publication etc.”

We tried to keep the documentation on the Docker usage simple and not redundant with the already extensive documentation available for the Galaxy Docker project (https://github.com/bgruening/docker-galaxy-stable). In the online documentation, we added more links in the documentation to this Docker documentation, especially with the questions the reviewer asked, and added a sentence to refer to this online documentation in the manuscript. To answer the question directly, it is possible to store, archive and share the entire /export folder of a Galaxy Docker image. This can then be easily shared, uploaded to Zenodo etc. and reused with any other Galaxy Docker container.

“Overall an excellent paper !”

Reviewer #2

“Some spelling errors:

line 56: blocking scientist -> blocking scientists line 124 an web-interface -> a web-interface line 135: visualization -> visualizations line 135: such Phyloviz -> such as Phyloviz line 157 Figure 2): we -> Figure2) and we line 175-176: We integrate then also a workflow -> We also integrated them in a workflow

in report (supp. material) targeted abundances may be not reflect -> targeted abundances may not reflect”

Thanks for reporting these mistakes, we addressed all of them in the revised version.

Further remarks:

1) “The title is a bit lacking in context. ASAIM is clearly dedicated only towards the taxonomic and functional analysis of metagenomic data (either from amplicon sequencing or from shotgun sequencing). It would be beneficial for the reader to deduce that from the title.”

ASaiM is a community starting point for all people interested in metagenomic research. During the last months other tools related to metagenomic assembly as MetaSPAdes or MEGAHIT and some tools for binning were added, partially by the community, but also on request from collaborators. The objectives of ASaiM is to offer a comprehensive and general workbench for microbiota analysis and thus we would like to have a slightly more general title. However, we changed the title slightly to: “ASaiM: a Galaxy-based framework to analyze microbiota data”

2) “It's not quite clear the innovative part of the platform. Besides collecting all those preexisting tools in an organized manner under Galaxy's umbrella what was the added contribution of ASAIM's team? Did you develop new wrapper/parser scripts for some/all of these tools in order to integrate them with Galaxy? What is the added value of the 3 new tools you developped? The GO slim term tool seems to be one of the final tools (purple) in your workflow (is that correct?). What about the other two for searching EBI and ENA databases? Are they part of one of the workflows or just additional standalone tools?”

The ASaiM team migrated 12 tools/suites of tools and their dependencies to Bioconda (e.g. HUMAnN2, MetaPhlAn2, GraPhlAN), integrated 16 suites (>100 tools) into Galaxy (e.g. HUMAn2 or QIIME with its around forty tools), i.e. developing the wrappers for these tools. We also checked and updated the wrappers of the existing tools. Moreover, several Galaxy datatypes, (interactive) training material and a visualization were developed and integrated into Galaxy. The 3 tools we developed were needed to close missing steps in workflows or to make it more convenient for users to access publicly available data. The GO slim tool is used to aggregate the gene family abundances into GO terms and is indeed one of the final steps in the workflow. The EBISearch and ENASearch tools are standalone tools to allow users to query ENA and EBI Metagenomics databases (data, metadata) and transfer to directly into Galaxy. They are not integrated into the one of our predefined workflows because the inputs of the workflows could be local data or data from external database such as ENA and we can not determine that before. To complement the tools and workflows, the ASaiM team created also documentation and tutorials.

3) ”The comparison between ASAIM and EBI analysis seems rather trivial. It's not a comparison of the two platforms rather than a comparison of the two different tools they are using (QIIME and Metaphlan). It would make much more sense a comparison between EBI's workflow run in the exact same way as an ASAIM/Galaxy workflow with the same tools.”

We would like to do this, but currently it is not possible to know the exact parameters which are used in the EBI Metagenomics workflow. This latter workflow is, unfortunately, currently a blackbox in contrary to ASaiM whose one of the objectives is to make microbiota research more transparent and reproducible.

4) “The same goes for functional analysis (where you mention comparison is not feasible). You just present results derived from two different methods with no comparable points.”

For the same reason as stated above we are very limited in what we can compare. Moreover. the functional information are extracted with two different types of information. EBI Metagenomics extracts the InterProScan gene families. In ASaiM, we extract with HUMAnN2 the UniRef gene families. It complexifies any comparisons.

5) “In line 200 the command you state docker run -d -p 8080:80 quay.io/bebatut/asaim

is different than the one stated in your webpage where the installation instructions are:

docker run -d -p 8080:80 quay.io/bebatut/asaim-framework

while the "asaim" command doesn't work (not authorized error) the "asaim-framework" seems to work”

We apologize for this mistake. We fixed the command mentioned in the manuscript to fit to the one in the instructions.

6) “In supplementary material report page 3 contains a table that is not well displayed”

Thanks for reporting this. We fixed the table.

7) “Installation was not succesful so actual testing of the tool was not possible. Installation in a new CentOS distribution (3.10.0-514) under a Virtuabox engine failed. It could be useful to mention in your docs how to install and start the docker engine before attempting to download the ASAIM package especially for those with little or no command line knowledge.”

As the installation of the Docker engine can vary between different operating systems and is changing over time we think the best way is to link to the upstream documentation under https://docs.docker.com/engine/installation. We also added a link to a video explaining how to use Kitematic for Galaxy Docker, for the non-linux users.

“At some point during the installation process there was an error saying:

"failed to register layer: ApplyLayer exit status 1 stdout: stderr: write /tool_deps/_conda/envs/__picrust@1.1.1/lib/python2.7/site-packages/mpi4py/MPI.so: no space left on device."

Not sure how that's possible with 34GB available free space. Does ASAIM include databases that take up more space than that? If that's the case you should probably include that in the Requirements section in your webpage and inform the reviewers as well in order for us to be able to succesfully install and properly test it.”

We apologize for this unfortunate experience. ASaiM includes numerous tools and reference databases for HUMAnN2 and MetaPhlAn2 and this increases the required disk space to 40GB. We forgot to mention this in our documentation and addressed this issue. In the meantime we are working hard to make this experience easier in the near future. The latest ASaiM Docker release already supports the CVMFS filesystem, with which we can easily mount in TB of reference data into every image. The data is then only downloaded if it get accessed by tools. We will extend this over the next releases.

Reviewer #3

“The manuscript describes an alternative workflow for the processing of shotgun metagenomics and metatranscriptomic data, called ASaiM. ASaiM integrates multiple tools for the analysis and manipulation of raw metagenomics and metatranscriptomic data, that are available, both as single tools and combined in multiple pipelines, within the Galaxy workflow and with a Docker and conda support. ASaiM comes with a very impressive documentation and it is of high importance in the metagenomics community, where most of the analyses are carried out using in-house scripts that, as pointed out by the authors, hinder reproducibility.

However, several other metagenomics pipelines are already available: MG-RAST and the EBI metagenomics pipeline, that the authors briefly discuss in the Introduction, but also MOCAT2, MetAMOS, and another Galaxy metagenomic pipeline. How does ASaiM compare within this wider ecosystem? MOCAT2, for instance, comes with a set of preset parameters, stored in a single file, that already improve reproducibility, and the EBI metagenomics pipeline clearly shows the software version (e.g., https://www.ebi.ac.uk/metagenomics/pipelines/3.0), allowing provenance.”

Provenance is way more than just the version of the used tool in a workflow. Every single parameter or the version of the used reference database can have a huge influence on the results. But even if the various webservers would allow for a complete provenance it’s hard or impossible to run those pipelines locally or on a local cluster. ASaiM is changing this by offering all tools of the different pipelines in one workbench, that can be deployed locally, on a cluster or in a cloud. The different pipelines can even be mixed if necessary, allowing for a unmatched flexibility and reproducibility. Moreover, ASaiM will ensure that the entire provenance is tracked and every single parameter, the exact version of the tools and input data is tracked and can be reproduced and compared. The reviewer mentioned MOCAT2. This command line tool is a great tool. However, it focuses only on metagenomic data (not for metataxonomic or metatranscriptomic data, as we would like) and its command-line use is a limitation for its use for all scientists working with microbiota data. We will work on integrating it into ASaiM. With EBI Metagenomics, the versions of software are available but not the parameters or the versions of databases used. For this reason, we did not set up any parameters in the workflow developed to reproduce the one on EBI Metagenomics. We think it is a big issue for reproducibility, as the parameters and the databases can have a big impact on the results.

“Also, the authors point out that the main problems in analysing metagenomics data are, first, the selection and configuration of the necessary tools, then the definition of the correct computational resources, and, finally, the definition of a correct analysis workflow. However, in this reviewer's opinion, ASaiM does not fully address these limitations. The authors implement about 25 tools for the processing of metagenomics data but give little explanation on the reasons these specific tools have been selected, or which tools should be used when multiple tools within the same class are available. Novices in the field would surely appreciate these pieces of information as a way to select the correct software for the problem at hand.”

Information about this was added in the documentation and in the tutorials we developed with the Galaxy Training Network. We follow the idea to offer a variety of different tools, even if they have overlapping functionality, to enable a lot of flexibility and freedom in data analysis. In this regard we want to offer easy access to a lot of different software. If a user needs guidance and the amount of tools is just overwhelming, we provide workflows for different use-cases and training material, in which we choose specialised tools and leave other out. However, we think the power of an analysis should be in the hand of the user and different steps in a workflow should/could be interchangeable.

“Regarding the workflows included in ASaiM, one is a reimplementation of the EBI workflow, one cannot be used for analysing metagenomic shotgun data, and only one is novel (that this reviewer supposes is the one called very generally "ASaiM"). This reviewer would suggest the authors to focus more on describing this novel workflow, and to remove all the references to QIIME and Mothur tools (or to 16S data analysis in general) since these are not able to analyse shotgun metagenomics data and may generate confusion.”

We think that ASaiM should be general toolkit for the analysis of microbiota data, not only for shotgun data. Microbiota analyses are usually not only focused on one type of analysis (metagenomics, metatranscriptomics, metataxonomics). We usually need to combine tools developed for different purposes to analyze our data. For example to compute abundance statistics such as alpha or beta diversity, we can apply the QIIME tools on the BIOM files generated by metagenomics tools such as MetaPhlAn. ASaiM is currently used in diverse microbiota projects (shotgun, amplicon and ITS data). We would like then to keep the mention of the QIIME and Mothur tools, and their workflow. We also added two workflows for metagenomic assembly (one using MEGAHIT and one using MetaSPAdes), including quality control, assembly and assembly checking (statistics, mapping and identification of potential assembly error signatures).

“For instance, it would be interesting to know how the workflow can be customised, whether default parameters are available and how they have been selected, and have more detailed and exhaustive information on time and computational requirements (and not only on two samples).”

We clarified the customization of workflows in the manuscript: “To assist in microbiota analyses, several default workflows are proposed and documented (tools, default parameters) in ASaiM. These workflows can be used as they are, customized either on the fly to tune the parameters or globally to change the tools, their order and their default parameters, or even used as subworkflows.”. We added more details in the documentation and also in the tutorials about the choices of default parameters for the tools. Exhaustive information on time and computational requirements are difficult to extract. They greatly depend on the input data. Currently for the shotgun workflow, the main time-consuming task is HUMAnN2 and its execution time is not linear with input size. We added a sentence in the manuscript to mention that. In general ASaiM is configured by default to run on normal personal computers, but because ASaiM is utilizing the Galaxy framework all tools and workflows can be easily configured to scale out and use entire clusters or other available compute resources. Here, we are referring to the upstream documentation of Galaxy or the Docker Galaxy project.

“Also, it is not clear what improvements are brought by ASaiM and what are due to the usage of Galaxy (reproducibility, provenance, being user-friendly), or of HUMAnN2 (ability to infer the taxonomic profiles up to the species level, availability of genes and pathways abundances tables). For instance, how the proposed 'functional and taxonomic combination analysis' block differs with that proposed within the HUMAnN2 pipeline?”

ASaiM is a collection of existing tools that are combined into a dedicated Galaxy instance. On top of these tools we have build workflows and training material. Thanks to Galaxy and Docker, ASaiM can be easily shipped, deployed, but also customized for anyone. The ASaiM team maintains the tools, updates them, integrates new tools (> 100), datatypes and visualizations and develops documentation and training to help researchers to deal with microbiota data. We clarified the manuscript in this direction. The “functional and taxonomic combination analysis” block is the Galaxy implementation of the HUMAnN2 pipeline, but inside a workflow to help its execution on many samples and after several pre-processing steps (quality control, sorting, MetaPhlAn2), without the need to care about the computational details. It is a turnkey solution.

“More in general, this reviewer's main concern regards the focus of the manuscript. Are the authors interested in presenting the Galaxy implementation of a variety of metagenomics tools? Or to present a novel reproducible pipeline for the analysis of metagenomics data? Are they interested in metagenomics or metagenetics (16S) analysis? In this reviewer's opinion, the manuscript would surely benefit in focusing on a single message, while additional features (such as the analysis of metagenetics data) should be only briefly mentioned.”

We are interesting in presenting ASaiM as an environment for people working on any type of microbiota data: a Galaxy implementation including a variety of microbiota related tools, workflow, documentation and training, whch is easy to distribute with its Docker image, for example for a publication of an analysis. We tried to make this message clearer in the manuscript, with for example a slightly different title “ASaiM: a Galaxy-based framework to analyze microbiota data”

“The manuscript includes some imprecision, with several concepts repeated multiple times, and would surely benefit from a proofreading by a native speaker:”

1. “Lines 40-43. Metagenomics and metatranscriptomics techniques do not allow to get insight into metabolic components, but only on the inferred functions of the micro-organisms present in one sample (as done, for instance, by HUMAnN2). To measure the metabolic components, one should use another approach, namely metametabolomics. It is also not clear what 'phylogenetic properties' are. Do the authors mean taxonomical profiles?”

We changed the sentence to clarify it: “These techniques are giving insight into taxonomic profiles and genomic components of microbial communities.”

1. “Line 44. The authors mention 'high variability'. What is the feature showing this 'high variability'?”

High variability is referring to the diversity of organisms in one sample, uneven sequencing depth of the different organisms and other things that makes metagenomic research hard. We changed the word to use “their complexity”.

1. “Line 52. Can the authors give examples of what they call 'computational resources specially for the metagenomics datasets'?”

We meant need for lot of memory and disk space, the use of cluster or cloud. They are not specific for metagenomic datasets, but probably highly required for metagenomics. We changed the sentence to: “They are command-line tools and may require extensive computational resources (memory, disk space)”.

1. “Line 140. What is a 'data reduction step'?”

A data reduction step is the reduction of the input data: removal of bad quality sequences and trimming, removal of duplicated sequences (dereplication), sorting of the sequences. We removed this term, to avoid confusion.

1. “This reviewer suggests removing the 'Installation and running section' and simply refers to the documentation, as done in other cases.”

We decided to have this section because it shows that using ASaiM is not really difficult and also to mention that tools and workflows can be added to any already existing Galaxy instance. We significantly shortened this section and referenced the documentation. Thanks for this recommendation.

While this revised version of the manuscript improves on the previously submitted one, this Reviewer believes that a few points still need to be addressed:

1. While this Reviewer agrees that ASaiM allows users to overcome "the difficulty to find, configure, use and combine the dedicated bioinformatics tools", it is still true that "to extract useful information, a sequenced microbiota sample has to be processed by sophisticated workflows with numerous successive bioinformatics steps", that "Each step may require execution of several tools or software", that "[tools] may require extensive computational resources (memory, disk space)", and, finally, that "selecting the best tools, configuring them to use the correct parameters and appropriate computational resources and combining them together in an analysis chain is a complex and error-prone process.". This Reviewer suggests reframing the manuscript either stressing on ASaiM's strengths compared to state-of-the-art tools (that is, in this Reviewer's opinion, saving the users from the hassle of installing all the pieces of software, and implementing a few well-known pipelines into Galaxy, an universally-acknowledged user-friendly platform), or clarifying, how ASaiM solves the issues raised above (that is, mostly, how i) ASaiM diminishes the memory/space requirements, ii) helps users in designing novel meaningful pipelines using the >100 tools included, and iii) helps users in setting meaningful parameters/resources in each of these steps). Following on this comment, the limitation of both QIIME and Mothur that is: "Designed for amplicon data, both QIIME and Mothur can not be directly applied to shotgun metagenomics data." is still not addressed by their ASaiM implementation and should, in this Reviewer's opinion, be removed.

2. In this Reviewer's opinion, the comparison between ASaiM and the EBI pipeline is irrelevant, since they use different tools (and it rather seems a comparison between these tools). If the authors cannot provide a fair comparison, this paragraph could, in this Reviewer's opinion, be removed without loss of information.

3. This Reviewer agrees that time and other computational requirements greatly depend on the input data, and thus suggests carrying on a benchmarking of all the implemented pipelines using multiple datasets, with different numbers of reads (many, as those belonging to the Hunan Metagenome Project, are freely available). This will help users in "selecting the best tools, configuring them to use the correct parameters and appropriate computational resources", and give them more useful information than that which can be extracted by only two datasets.

4. Minor comment: since there is no agreement yet on some of the terms used, it may be worth using 16S rRNA marker gene sequencing or amplicon sequencing, instead of metataxonomic, and whole metagenomic shotgun sequencing, instead of simply metagenomics.

Level of interest Please indicate how interesting you found the manuscript:
An article whose findings are important to those with closely related research interests

Quality of written English Please indicate the quality of language in the manuscript:
Acceptable

Declaration of competing interests Please complete a declaration of competing interests, considering the following questions: Have you in the past five years received reimbursements, fees, funding, or salary from an organisation that may in any way gain or lose financially from the publication of this manuscript, either now or in the future? Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this manuscript, either now or in the future? Do you hold or are you currently applying for any patents relating to the content of the manuscript? Have you received reimbursements, fees, funding, or salary from an organization that holds or has applied for patents relating to the content of the manuscript? Do you have any other financial competing interests? Do you have any non-financial competing interests in relation to this paper? If you can answer no to all of the above, write 'I declare that I have no competing interests' below. If your reply is yes to any, please give details below.
I declare that I have no competing interests.

I agree to the open peer review policy of the journal. I understand that my name will be included on my report to the authors and, if the manuscript is accepted for publication, my named report including any attachments I upload will be posted on the website along with the authors' responses. I agree for my report to be made available under an Open Access Creative Commons CC-BY license (http://creativecommons.org/licenses/by/4.0/). I understand that any comments which I do not wish to be included in my named report can be included as confidential comments to the editors, which will not be published.
I agree to the open peer review policy of the journal.

Authors' response to reviews: Reviewer #2

Please make sure to double check the Figures before publication. Figure 3 seems to have its title overlap with some other text.

We are sorry for the inconvenience, but can not see this locally. We will wait for the proofs and make sure this does not end up in the final PDF.

Reviewer #3

While this revised version of the manuscript improves on the previously submitted one, this Reviewer believes that a few points still need to be addressed:

1. While this Reviewer agrees that ASaiM allows users to overcome "the difficulty to find, configure, use and combine the dedicated bioinformatics tools", it is still true that "to extract useful information, a sequenced microbiota sample has to be processed by sophisticated workflows with numerous successive bioinformatics steps", that "Each step may require execution of several tools or software", that "[tools] may require extensive computational resources (memory, disk space)", and, finally, that "selecting the best tools, configuring them to use the correct parameters and appropriate computational resources and combining them together in an analysis chain is a complex and error-prone process.". This Reviewer suggests reframing the manuscript either stressing on ASaiM's strengths compared to state-of-the-art tools (that is, in this Reviewer's opinion, saving the users from the hassle of installing all the pieces of software, and implementing a few well-known pipelines into Galaxy, an universally-acknowledged user-friendly platform), or clarifying, how ASaiM solves the issues raised above (that is, mostly, how i) ASaiM diminishes the memory/space requirements, ii) helps users in designing novel meaningful pipelines using the >100 tools included, and iii) helps users in setting meaningful parameters/resources in each of these steps). Following on this comment, the limitation of both QIIME and Mothur that is: "Designed for amplicon data, both QIIME and Mothur can not be directly applied to shotgun metagenomics data." is still not addressed by their ASaiM implementation and should, in this Reviewer's opinion, be removed.

The authors understand the first point of the reviewer and tried to clarify in the manuscript how ASaiM solves the raised issues. In the introduction of the workflow section, the authors add sentences to show how ASaiM helps users in setting meaningful parameters for tools and also in designing novel meaningful workflows. In the conclusion, the authors added few words to insist on the automated hassle of tool installation. Galaxy via ASaiM will not address the memory limitations, but Galaxy will efficiently schedule jobs as well as manage the memory usage. This information has also been added in the manuscript. For Mothur and QIIME related question, ASaiM offered tools and workflows for amplicon or metataxomic data using QIIME and Mothur, but also for shotgun metagenomics data (using MetaPhlAn2 and HUMAnN2). ASaiM is not only then focused on amplicon data as QIIME and Mothur are.

1. In this Reviewer's opinion, the comparison between ASaiM and the EBI pipeline is irrelevant, since they use different tools (and it rather seems a comparison between these tools). If the authors cannot provide a fair comparison, this paragraph could, in this Reviewer's opinion, be removed without loss of information.

The idea of the comparison between ASaiM and EBI metagenomics was to demonstrate the limitation of one approach against the other on the analysis of shotgun metagenomic data. We are not benchmarking the tools, just trying to illustrate the possibilities of ASaiM.

1. This Reviewer agrees that time and other computational requirements greatly depend on the input data, and thus suggests carrying on a benchmarking of all the implemented pipelines using multiple datasets, with different numbers of reads (many, as those belonging to the Hunan Metagenome Project, are freely available). This will help users in "selecting the best tools, configuring them to use the correct parameters and appropriate computational resources", and give them more useful information than that which can be extracted by only two datasets.

Such general benchmarking would be interesting and we are working currently together with other researchers to establish a general benchmarking, as mentioned by the reviewer. Using the information of the benchmarking, we would like to build an environment where users could be helped with tool selection and configuration and jobs/workflows automatically tweaks in Galaxy. We feel that this is out of scope for the manuscript but we are working on this as a more general framework, probably not only for metagenomics.

1. Minor comment: since there is no agreement yet on some of the terms used, it may be worth using 16S rRNA marker gene sequencing or amplicon sequencing, instead of metataxonomic, and whole metagenomic shotgun sequencing, instead of simply metagenomics.

The authors agree that there is a confusion in vocabulary used in the field of microbial community analysis. Marchesi & Ravel in their 2015 paper (Microbiome: (https://microbiomejournal.biomedcentral.com/articles/10.1186/s40168-015-0094-5) tried to establish a consensus vocabulary. To support this initiative, the authors decided to use the terms and definitions given in this paper. We hope this makes our paper more readable in the long run and supports the initiative started by Marchesi & Ravel.