The authors report on a metagenome and metatranscriptome for a biogas-producing microbial community. The authors provided a Docker container as an aid for reproducing the metagenome assembly which was tested in my review.

Following the instructions provided by the authors [1], I was able to use the Docker container to execute the metagenome assembly workflow which the authors have implemented as a Makefile. The Docker container was tested using an AWS r3.8xlarge instance with 32 cores, 244 GB RAM and 100 GB storage.

The contigs produced by the Docker container on my AWS instance showed that they were similar but not identical to those obtained by the authors. This was confirmed with a blast all vs all comparison of the contig sequences. These small differences in my contig sequences affected the gene prediction analyses further along the metagenome workflow.

After discussing the differences between contig files with the authors, we believe that the small differences in contig sequences are caused by the non-deterministic nature of the Ray metagenome assembler. A multi-threaded run of Ray can cause different seed reads to be selected each time for building contigs [2]. This appears to explain why the gene prediction results I obtained are different to those of the authors. I believe this is a plausible reason for why my results differed from those of the authors and, as such, their Docker container reproduces their metagenome assembly process as reported in the manuscript.

Minor Essential Revisions

There are 2 typos in the README.md file [3] in the Docker container section; "wokspace" should be "workspace".

It would be useful to let users know how long it takes for the Makefile to complete its execution on a r3.8xlarge AWS instance.

1. https://github.com/metagenomics/2015-biogas-cebitec
2. http://sourceforge.net/p/denovoassembler/mailman/message/31936187/
3. https://github.com/metagenomics/2015-biogas-cebitec/blob/master/README.md 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 work for GigaScience as a bioinformatician.

Authors' response to reviewer: (http://www.gigasciencejournal.com/imedia/2020536111178956_comment.pdf)