blog

Dr. Hans Dijkstra is a researcher studying molecular evolution at the Institute for Comprehensive Medical Science at the Fujita Health University in Japan. When reading a recent publication in Science he noticed several flaws in a study on a long noncoding RNA. Dijkstra explains why he started a discussion on this report on Inc-DC RNA and the use of RNAi experiments and why he chose to do this via an F1000Research correspondence article co-authored by Keith Ballingall from the Moredun Research Institute in Scotland.

F1000Research: Can you briefly tell us what your correspondence article is about?

We claim that a recent article in Science by Wang et al. has flaws regarding the presentation of its long noncoding RNA (lncRNA) functional model. We also feel that the authors did not perform sufficiently exhaustive RNA interference (RNAi) control experiments for allowing robust conclusions. Wang et al. claim that the 3’-end of a human lncRNA expressed by developing dendritic cells (DC) binds to transcription factor STAT3, and that by preventing the de-activation of phosphorylated STAT3 protein this binding stimulates dendritic cell development. Although in our opinion Wang et al. made their conclusions somewhat prematurely, in humans they may be true. However, Wang et al. also used mice to confirm part of their data, yet failed mention that in rodents the orthologue of the human transcript which they call “lnc-DC” had been already coined “Wdnm1-like” by another study. This rodent orthologue has established protein coding capacity, and is expressed by adipocytes upon adipogenesis. Intriguingly, our genetic analyses suggest that among extant eutherian mammals, humans may be the only species without Wdnm1-like protein. Figure 1 summarizes the above, and emphasizes the interesting question as to how the various reported Wdnm1-like functions can be reconciled into a single evolutionary model.

F1000Research: What impact do you hope this discussion will have on the research community working on Inc-DC and Wdnm1-like and people using RNAi experiments?

Well, I hope this raises a lot of interest in Wdnm1-like protein. Adipogenesis and dendritic cell development are important processes, also from a medical/commercial point of view. As we described in the article, there appears to be a reasonable chance that Wdnm1-like protein from for example gorilla can still function on human cells.
It is also interesting to note that when a topic (like long noncoding RNA) is becoming a hype, it seems to bypass the critical consideration that new discoveries are usually subjected to. We feel it is very important to draw attention to the fact that a large number of co-authors, editors at Science and referees have missed the fact that murine Wdnm1-like does encode a protein. The Science referees probably were not aware of this protein coding capacity and interpreted the RNAi experiments that Wang et al. performed in mice as a statistical confirmation for the presented human long noncoding RNA functional model.
Regarding the technical discussion on RNAi experiments, as to when such experiments are sufficiently exhaustive for allowing robust conclusions, I am very happy with the expert referee report by Dr. Burchard. I hope more RNAi experts will state their opinion. Ideally, from a consensus among experts the journals may set up some minimal requests for the controls needed in RNAi experiments, especially if the RNAi results are so critical for the major conclusions.

F1000Research: Why did you decide to publish a correspondence article in F1000Research rather than comment on the article in Science?

Initially we sent our article to Science, since we feel that would have been the most appropriate place for a discussion on, in our opinion, major flaws in one of their recent publications. However, the Science editors only allowed us to place a message online underneath the publication by Wang et al. We feel that Science does not particularly draw attention to this type of comments, also regarding the lay-out of their site. In addition, the number of words that can be used in these comments is very limited. We thus decided to publish in F1000Research instead, since it encourages a full-sized and open debate, which is readily detected by the research community because it is indexed in PubMed. For the sake of honest debate, we also find the co-publication of the non-anonymous referee reports highly advantageous, since it should help the readers come to well-balanced conclusions. We did leave a short online message in Science underneath the article by Wang et al., referring to our discussion in F1000Research.

F1000Research: How did you experience the peer review process at F1000Research?

I consider both the report by Dr. Smas and Dr. Ren, and the report by Dr. Burchard, excellent. Compared to reviews that I generally receive from referees at other journals, these are better written. This may have to do with the respective referees, but probably also has to do with the fact that these reports are published online along with the referees’ names. However, I was disappointed with how difficult it was to find referees. For example, the Wang et al. group has not commented so far, and a relatively large number of invited experts were unavailable to review. I wonder if it would have been easier to find referees if the reports were anonymous and not published online. [Editor’s note: finding sufficient referees during the summer period is always a challenge for journals. However, more and more journals are moving towards open-peer review].

F1000Research: Would you submit to F1000Research again in the future?

In a similar situation, when I like to initiate a discussion, definitely. I was very pleased with the submission and editorial process and with the resulting article. It gives a refreshing and clean feeling. Of course the current lack of impact factor of the journal is a luxury which I can’t afford for all my publications. However, when F1000Research has an impact factor, I will definitely consider publishing a Research Article. I have read a discussion on ResearchGate where people seemed very pleased with the publication of their Research Articles in F1000Research.