We have waited a week to respond in case the promised more complete response was about to appear (and we look forward to reading it), but we shouldn’t let this issue drag on.

You say that you welcome robust tests of your reported findings, and that is exactly what we seek. We have shown a lack of evidence of iridovirus and nosema in two controls and one collapsing/collapsed, despite your reporting their presence in all three of these samples. Having demonstrated that the small subset of raw data that was eventually released was unreliably interpreted, it is now even clearer that the rest of the data needs to be re-evaluated.

Hence, we renew our request to make available to the community the remaining raw data that you report shows evidence of the presence of iridovirus and nosema. All that is required is the .raw mass spectrometry files, which can be freely uploaded to a site such as Tranche, in the way we did in our manuscript, and as required to publish proteomic results in most journals publishing this type of data. The ECBC report here (http://www.dtic.mil/cgi-b...) contains sufficient information, in combination with the .raw files, for the proteomic community to evaluate these data and database search results.

If the scientific process is allowed to take its course, these data will speak for themselves.

Best regards,

Giselle Knudsen and Robert Chalkley

Dr. Firth,
I appreciate your enthusiasm and that of your co-authors, many of whom I consider friends. Nevertheless, concerns over this work have been raised since publication from the angle of
1) biological improbability (the 'presence' of more Nosema species than have been found across all insects previously, the claimed presence of hundreds of other microbes in a host known for having a relatively simple microbiome, and claims of a ubiquitous DNA virus not yet seen by other methods despite strong evidence for other DNA and RNA viruses in bees),
2) stoichiometry (i.e., equal representation of all peptides in the focal species, a putative iridovirus)
3) imprecise methods for each step of the proteomic component from sampling to database searching, despite wide evidence that these steps can give precisely the erratic results many now believe to define this screening (see Foster Mol. Cell Proteomics 2011 and the above paper by Knudsen and Chalkley).
and
4) a persistent promise since last fall that the underlying data will be revealed as a defense against these and other strong critiques.

It is not profitable to critique other scientists and you must realize by now that the main drive to do so is because many of us have ‘seen this movie’ wherein hubris and reluctance to question ones own results can put science in a tailspin. Knudsen and Chalkley worked hard to unearth and reanalyze three samples that were claimed to contain IIV and in each case found a more plausible alternative. This result was found to be convincing by reviewers and PlosONe editors. I would urge you in weeks, not months, to release further raw data from this study if you wish to reverse a growing sense that the main results were a sad, if presumably honest, mistake. Your coauthors and the community deserve a swift resolution of this, and no-one has benefitted from nine months of silence.
Respectfully,
Jay