The authors do a nice job of describing the fMRI and accompanying data featured in their article "Single subject fMRI test–retest reliability metrics and confounding factors" in the journal NeuroImage. However, while the abstract refers to "Findings" concerning the reliability of these data, no summary of these results is presented. Although this was the topic of a related article it might be advisable to have a section briefly describing just how reliable these data are and what that might imply for their usage by others. Of course, readers can be referred to the original paper for additional details but having a summary ties the whole thing together as a stand alone article for GigaScience.

The authors claim that the data sets "provides an opportunity to study the fusion between fMRI and DTI datasets" however I'm not certain that this is a major plus of these particular data or this article, per se, since image registration, parcellation, and multimodal "fusion" are so often performed in the literature simply as part of the process of data processing. In other words, combining functional and structural brain imaging data and even DTI data is not new nor would it be advanced significantly through the use of these data. What might more convincing as an argument is that the various calculations performed on the data indicate them to have high reliability, likely to be highly replicable, have a sufficient level of statistical power, etc. These may be of interest to machine learning theorists, pattern recognition specialists, and others working to find novel ways to tease apart the spectrum of effects underlying BOLD related signal change during cognitive stimulation.

The authors also suggest that "To our knowledge there are no other publicly available test-retest dataset which provide five different task-based fMRI paradigms". The authors might wish to examine the contents of efforts such as the fMRI Data Center (fmridc.org) and the OpenfMRI project (openfmri.org) which contain several data set from peer-reviewed studies of cognitive neuroimaging using fMRI which present multiple task paradigm manipulations. The authors should cite these resources and any published research articles which describe them.

However, it is nice to see authors sharing their data with the community and making the effort to do so in a well organized and efficient manner. With the above suggested improvements, these data will certainly find utility in the community.

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: No competing interests to declare.

Thank you for addressing my concerns. I have nothing further to add and recommend acceptance of the manuscript as is.

Level of interest: An article whose findings are important to those with closely related research interests

Quality of written English: Acceptable

Statistical review: No, the manuscript does not need to be seen by a statistician.

Declaration of competing interests: No.