Review for "A reference bacterial genome dataset generated on the MinION™ portable single-molecule nanopore sequencer"

Completed on 4 Oct 2014 by Keith Robison .

License: http://creativecommons.org/licenses/by/3.0/

Login to endorse this review.


Comments to author

This report is primarily the summary of a dataset which has been made available, with some simple analyses of its bulk properties. The methodology is soundly described. This report will greatly advance the rapidly shifting field of MinION analysis methodology by providing a body of data for community analysis. The authors have done an admirable job of specifying the precise versions of the various components of the MinION platform, which will be critical for later analysts who may use different versions of this rapidly changing system.

Major Compulsory Revisions
The Methods section does not appear to mention spiking in the provided control sequence (a fragment of phage lambda), which is part of the standard MinION library preparation protocol. Because MG1655 is lambda+, it is difficult to determine from the data whether or not this control was used; it should be made explicit in the Methods.

Figure 2, if retained in its current form (see below), must have the color key enlarged -- it is very difficult for this particular reviewer, due to a partial color vision deficit, to distinguish them. Replacing the superfluous box in the key around each color, and instead making the colored spot as large as that box, would remedy this.

Minor Essential Revisions
The match parameter for LAST was presumably left at the default of 1, but it would be useful to newcomers to have this stated explicitly typo: "smaple"

Discretionary Revisions
Figure 2 is probably the most significant bit of analysis, presenting the sequencing accuracy based on alignments to reference as a 2x2 grid (trellis) of plots. This is a powerful idea, but I think the authors should explore an even greater trellising. Each plot has four different types of data (Template, Complement, Normal 2D and Full 2D) distinguished by color. Given the density of the datapoints, seeing the exact boundaries of these 4 is difficult, particularly for this reviewer. Further trellising on the 4 data types would certainly make them distinguishable, assist in showing the different distributions for each type & avoid using color for contrast.

Given the novelty of the MinION system to much of the scientific community, a diagram indicating the library fragment structure and associated proteins would be a useful addition.

Level of interest An article of outstanding merit and interest 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