Reader Comments

Post a new comment on this article

Reanalysis of the full dataset indicates no IIV peptides

Posted by Apis_mellifera on 26 Sep 2011 at 18:03 GMT

First of all, thank you to the authors for releasing what is apparently the full set of raw data for this study at Tranche. I have gone through the same exercise with the authors' own data as I had done previously with another dataset, as well as the re-analysis done by Chalkley and Knudsen. In short, if honey bee AND viral proteins are considered in the database search then there is no confidently-identified iridovirus proteins identified at all. Conversely, if one does not consider the honey bee proteins in the search then there are several very weak identifications of iridovirus proteins.

Thus, reanalysis of this data supports the authors' own findings that IIV DNA cannot be amplified in these samples. The reason, now clear, is that there simply never was any iridovirus in the samples to begin with.

Leonard Foster

No competing interests declared.

RE: Reanalysis of the full dataset indicates no IIV peptides

chalkley replied to Apis_mellifera on 03 Oct 2011 at 23:18 GMT

They have now uploaded 68 raw data files from their study to Tranche. It can be accessed at the following page:
We have searched all of this data, and it appears that the three samples analyzed in our June PLoS One paper were representative of the rest of the data; i.e. there is no evidence for iridescent iridovirus in any of these samples. Indeed, there isn’t significant evidence of any viral protein in these samples. The main species identified in the samples were Apis Melifera (as expected), then Drosophila Melanogaster. Mosquito, wasp, bumble bee, moth and ant were all reported (along with human keratin), but we predict that some of these species identifications may not be exactly correct as we suspect the real species are probably not in the database, so these are the closest relations available. Hopefully, if the original authors are intending to further analyze this data, they will include at least these species in the database that they query.
In conclusion, our results independently agree with those reported by Leonard Foster that there is no evidence for iridoviral infection from these samples.

Competing interests declared: I authored a manuscript indicating a lack of evidence for iridoviral infection after analyzing three of the samples described in this manuscript.