Content of review 1, reviewed on July 29, 2013
Basic reporting
The paper meets the requirements
Experimental design
The paper meets the requirements
Validity of the findings
The paper meets the requirements
Comments for the author
The authors describe a number of tools and tool wrappers that have been integrated into Galaxy, and provide a use-case in molecular plant pathology
There could be more mention of alternatives to Galaxy, e.g. Taverna and Anvaya
Whilst MIRA has been integrated, no mention is used of the memory requirements - many are reluctant to integrate assemblers into their Galaxy instances for fear that several large memory jobs are launched by users
On page 5, two workflows are mentioned that are essentially identical, except one uses GetOrfs for gene finding and the second uses Augustus and Glimmer3. Doesn't the second workflow make the first redundant? Why include the first?
On page 6, technically I feel orthology should be the basis for transferring functional information, not sequence similarity. Similarly on page 7, isn't it more standard to use reciprocal best hit to define orthologues before transferring annotation?
Bottom of page 7, GetOrfs is used again - why not use the aforementioned gene predictors?
Was any attempt made to wrap the InterProScan web-service (rather than standalone)?
Top of page 8, I am curious whether the SignalP licence allows for it to be integrated into a public Galaxy?
The RXLR prediction tools: as I understand it, the authors have implemented several published methods for RXLR motif prediction, and released these into the Galaxy tool shed. Does this paper serve as notice of their publication? Has any testing been done on these implementations to demonstrate their accuracy and efficacy?
Overall the paper is well written and should be published. The above suggestions can be dealt with by adding text to various parts of the manuscript and do not represent a large body of work, therefore I recommend minor revisions
Mick Watson