Dear Readers,
We apologize to the readership for the decision by PLoS regarding our manuscript. This was based on concerns related to panels in some of the figures that appear to have vertical lines between the lanes. We had initially agreed with the decision because after more than 10 years, as well as changes in equipment used for scanning we were unable to identify some of the original images from the images in question. This would have deterred any questions as to their authenticity. However, we also stated that we stand behind our conclusions in the manuscript and believe that these studies are solid. Additionally, we did eventually find some images in one the notebooks which we sent to PLoS before the retraction but they were not accepted. We should also state that we do not agree with the retraction at this time because we believe that the data we provided, and the overall body of work is a substantial contribution to the field.

Moreover, we can state that vertical lines of concern, can and do certainly occur on scans of blots between lanes. In addition, we also provided an example of how this could happen using another blot that had vertical lines and showed the original blot to prove it was authentic. The lab-mate who performed these experiments was not able to find the original scanned images in question after months of searching but we did find the images in the notebook. Further, we provided Coomassie images of experiments with multiple repeated inductions of Gemin3 which demonstrated that they were repeated multiple times, as well as one of the original scanned images for the EBNA3C GST fusion polypeptides. In total, the manuscript provides a large body of evidence to support our overall conclusions and so we stand by our original conclusions in this manuscript.

Therefore, based on our findings of the Coomassie gel photos in the notebook, and the primary data submitted we no longer agree with the decision as we believe that we have sufficiently provided all data that was questioned. We should also add that we recently went back and again expressed Gemin3, and completed the pull-down of EBNA3C. The results were as expected, and was clearly reproduced once again in our hands. Therefore, our data is correct and reproducible. The data in question were also reviewed by an independent group of faculty who concluded that there were no concerns with data manipulations. We would be happy to directly review our data with investigators who have an interest in our work and would like to discuss further.

We hope that this clarifies our position and that readers understand our explanation of the data. Also, that these experiments were done more than 10 years ago with scanners that were 15-20 years old without the resolution that one expects to have today.

Best,

Erle and Qiliang
-----------------------------------------------------------------------------------------------------------------------------------------
More specific response to the editor
Query #11: The authors have provided individual level data underlying the graphs presented in Figs 3, 4, 5, and 6. They were unable to provide the original uncropped blots underlying the Fig. 1B WB:E3C panel, the Fig. 3D S35-Germin3 panel, the Fig. 5B S35-p53 panel, the Fig. 5C WB:Germin3 panel, and the Fig. 6B WB:Germin3 panel. Underlying blots were provided for the Fig 1B WB:Germin3 panel and the Fig. 4B WB:Germin3 panel, but these blots were of too low resolution to confirm the published results.

Response #11: During assembly of a couple panels we inadvertently used the incorrect image (i.e. duplicated in Fig.3C and 3D). However, we have now provided the correct and original images in response to the questions above, as well as provided very clear logical explanation as to why these lines can appear on images of the blots as can be seen when the images are enlarged and adjusted. Furthermore, we provided images of duplicated experiments showing that the experiments were reproducible and strongly supported our conclusions. None of these concerns affects the original conclusion and we have further expressed Gemin3 and showed that in fact it binds to EBNA3C in our recent experiments from a studies few months ago. Therefore, our work is reproducible and can be reproduced if anyone wants to repeat our work.

The majority of our data presented in this article were done more than a decade ago, and some image-capture or analysis software were not able to provide high resolution images and storage as today. Therefore, this has led to these queries and resulted in some of the images appearing to have these horizontal and vertical discontinuities surrounding lanes or between lanes as in Queries to Figs 3, 4, 5, and 6. We believe we have now fully addressed the speculation of discontinuous lines by the same problem appearing in images for example of Fig.4B WB:Gemin3 and Fig.5A S35-Gemin3, while their original uncropped images were clearly presented. In addition, when we changed the contrast one can clearly see that these lines are integral to the images in question and are not manipulated.

Although we could not find the original uncropped images for Fig.1B WB:E3C and WB:Gemin3, Fig.3D S35-Germin3, Fig. 5B S35-p53, Fig. 5C WB:Germin3, and Fig. 6B WB:Gemin3 panels in our scanned document files from more than 10 years, we were able to demonstrate why these may look discontinuous. Many changes in scanning equipment and storage of files transferred over the years has led to difficulties in recovering some of our images to date. However, and extremely important we did find some images containing the same results from repeat experiments (i.e. Fig.1B, WB:Gemin3, and Fig.5C WB:Gemin3, respectively). To further demonstrate to the readers that these discontinuities between lanes is due to the contrast and level of resolution of the images, we also provided the images in question with different exposures to the editors, particularly with increased contrast showing darker background which clarifies these concerns.

Taken together, we believe that our data presented in the paper fully support the conclusion. We did inadvertently included a couple incorrect panels during assemble and upload of the figures, and we sincerely apologize for any confusion this may have caused. We provided these corrected panels and have now repeated some of the key concerns including the expression of Gemin3 and its binding to EBNA3C which was reproduced. We requested approval to make corrections to these errors online, however, these were not acceptable to the editors and led to retraction of the manuscript.

For investigators in the field, who have a genuine interest in this study and would like to discuss the data and manuscript we invite you to reach out directly. We would be happy to discuss this with you openly as we believe and stand behind our work without waver. We believe that the work is solid and that it adds to the body of work we have spent decades building in the field.