When do you get it? A Nature research paper is worth nothing, if…

… it’s lacking the correct protocol! Yet another scandal on research papers shakes the scientific community. In the wake of the debate, the FAZ (Frankfurter Allgemeine Zeitung, for many the most relevant All-German newspaper), prints an interview with Rudolf Jaenisch, stem cell researcher at MIT, Boston, responding to the research paper by Dr. Haruko Obokata et al, claiming the production of so-called STAP cells.

The interview reveals insights into the whole story, which make it even more shocking and is pointing towards a general problem of the infrastructure within scientific communication and the publication of research papers: There is no integrity to the whole story, the processes and the raw data behind it.

A research paper example gone wrong

Okobota stated in a press conference, that she did the critical experiments 200 times, but would only state how exactly the experiments were done after the publication of her research paper. I’d like to agree to Rudolf Jaenisch’s comment: “That’s kindergarten, isn’t it?”

Besides, the protocol in the article was clearly not the correct one. Honestly: Everybody who is working in lab research knows that the methods described in research papers are worth nothing, details are lacking and there is no requirement for the authors to explain their procedures presented in the research paper at hand step-by-stepand easy to understand for everybody.

But back to the STAP-story: According to the FAZ article, the “correct protocol” was communicated via email to people that wanted to reproduce the findings of the research paper. Actually, four different protocols were sent around, as Rudolf Jaenisch states it. All of them didn’t work. No one was able to reproduce the findings of Okobota et al.

Welcome intransparency within scientific research papers?

Considering the latest scandal around Okobota’s research paper, several question come to mind: Why can’t a journal like Nature demand from researchers to provide the research paper, the correct protocols and all the raw data linked to it? And if there are so many versions of the protocol around, why isn’t it required to have a full audit trail of all the previous versions and a clear statement which version of the research paper let to which results?

In the hunt for a high impact story, it seems that both researchers and publishers embrace intransparency within research papers, backed by various excuses circulating around. Researchers often talk about the fear of getting scooped, but in truth it’s the fear of mistakes and failures in their research papers being discovered. It’s all about the chaos of too many and too careless documented procedures, observations and data being published.

In turn, publishers welcome intransparency in research papers, because otherwise they wouldn’t be able to make up great stories. Great stories are rather those that include a great vision and show the “impossible made possible”, and not those that include tedious and carefully documented research.

Getting published vs clean records in research papers

A fellow researcher and user of labfolder once stated: “Well, at institution X, they’re not really interested in well documented experiments, full audit trail etc. They want to write their Nature research paper and then it’s actually better to not have good records at hand.”

This was just recently and that’s why the STAP story was so personally shocking to me. It seems that the world of research is drifting apart into two different directions: The “boring” but reliable research sector and the “exciting” but highly not-applicable research corner.

This development is wrong. In my opinion, the solution should be: Everything that is connected to a scientific research paper should be interlinked at the time of publication and made open to the public. This includes:

  • all protocols referenced, with detailed description which procedures were changed,
  • all own protocols in great detail and in a step-by-step fashion
  • every version of protocols, with an exact statement which raw data resulted from which version of the research paper
  • all raw data and a full description of all analyses and which figures resulted from it,
  • all persons involved in the experiments: Who did what when for the research paper?

More transparency in research papers

Sure, within the process of research papers this would demand a more elaborate documentation. But modern times also provide the technology to link heterogenous data and formats to achieve a much greater transparency. At labfolder, we’re trying to give science a tool that enables researchers to do this.

However, it’s really time for publishers to understand that the published research paper alone is not sufficient. You’re not a good filter for more and more data, if you are not also the guardians of good laboratory practice, a quality assurance system for research.

 

Leave a Reply

Your email address will not be published. Required fields are marked *