General comments:

The manuscript entitled “Multi-platform miRNA profiling of FFPE samples from Hepatoblastoma patients” describes high-throughput miRNA expression data using three independent platforms: 1) Next-generation sequencing (NGS), 2) Affymetrix microarray, and 3) NanoString. Overall the manuscript is clear, well-written, and presents valuable data profiling an exceedingly rare cancer, however, the description of the data is similar to that presented in the authors’ referenced prior publication in Scientific Reports (PMID: 26039282). There are several questions that should be addressed for this manuscript to be appropriate for publication, the most important being whether value can be added over the previously published paper.

Major Compulsory Revisions:

1) It appears that all of the data were previously published in Scientific Reports, and it is not clear what, if any, differences or additions the current manuscript adds. Are there any additional samples that were recently run on any of the platforms that could be added? Did the authors attempt to integrate and/or validate their data with any other publicly available hepatoblastoma datasets, and perform an appropriate batch correction (e.g. GSE21085)? Can the authors comment further on future use of these datasets? As the Scientific Reports paper was a technical comparison, one could envision much more analysis of the clinical and biological characteristics of hepatoblastoma using these data. It would be helpful to describe these analyses.
2) The description of the data files in the “Availability, requirements and use” section states that the microarray data contain all miRNAs from miRBase 17, however the processed files for the NanoString data are reported as counts for each miRNA in miRBase 18. The NGS data were mapped to miRBase 17. For the purposes of cross-platform comparisons it would be useful to use a consistent mapping reference (i.e. one where miRNAs of a given mature sequence have an identical name). It would also be helpful to generate a table stating which miRNAs (with a consistent mapping reference) are detectable by each platform.
3) In the manuscript text the authors comment on the clinical outcome of one patient with a unique miRNA expression profile, however, no clinical data are presented in Table 1. It would be useful if the authors were able to provide any available clinical data (e.g. relapse-free survival, overall survival, treatment regimen, chemotherapy response, any genetic aberrations in addition to beta-catenin, etc.) to accompany the miRNA expression data. This should be in addition to the data already presented in Supplementary Table S1 from the Scientific Reports paper.
4) In the prior platform comparison paper, various technical controls and assay quality metrics are reported, and are referred to in the current manuscript. It would be helpful to present some of these data for each platform for each of the profiled samples (e.g. number of total reads, % reads mapping to miRBase, etc.).

Minor Essential Revisions:

1) Since FFPE tissues are often highly degraded, it would be useful to make a short point in the text regarding the minimum amount and quality of RNA needed to run these assays.
2) Several typographical and/or grammatical errors were noted in the manuscript text that should be corrected.
3) In processing the datasets software packages should be noted in Figure 1.
4) The title of the manuscript should avoid using abbreviations.
5) The references are not formatted in the style of the journal.

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/5963146581888086_comment.pdf)

The reviewed version of the manuscript can be seen here:

All revised versions are also available:
Draft -