Bruce Beutler, an immunologist at the University of Texas and winner of the 2011 Nobel Prize in Physiology or Medicine, was the most senior of several authors to publish the paper ‘MAVS, cGAS, and endogenous retroviruses in T-independent B cell responses’ in the journal Science in 2014.
The research in the paper states that virus-like elements in the human genome play a crucial role in the immune system’s response to pathogens.
After trying to replicate the paper’s results but obtaining contrary data, however, Beutler contacted Science to say that his latest experiments weakened his confidence in original results.
He and several co-authors of the paper requested a retraction of the original paper, but the process was drawn out when two of the co-authors stood by their original results.
Ming Zeng, the lead author, and Xiaolei Shi, a co-author, maintain that their original experiment is valid, and that the different finding was due to a problem with the design of the reproduction experiment.
Taking onboard the arguments from both sides, Science decided to retract the paper but, according to webofknowledge.com, the paper has already been cited 50 times. Therein lies the problem.
Modern science is one of the most sophisticated bouts of teamwork and determination imaginable. It involves hard work, incredible patience, and intense collaboration. Most notably, science constantly produces groundwork on which future work is based.
Imagine science as a card tower; the bottom layer of cards needs to be laid before the next layer can be built and stand without falling – has anyone ever made a successful card tower by starting from the top and working down?
This is where retractions become such a problem; imagine you’ve built a card tower, and then suddenly someone rips out one of the lower cards – all the cards above it may fall, and the time and effort to build the tower has been wasted. So too, in the case of a retraction, many or all of the papers that cited the retracted paper can be called into question. For the affected papers, each paper must then be carefully analysed – taking time away from other work to do so.
To combat the risk of publishing a paper that could later be retracted, papers are peer reviewed, although not always with high scrutiny.
In other cases, the experimental data could be a fluke; if all the data pointed towards a conclusion, even just by chance, how could someone tell? Even the most eloquent paper, with solid controls and data behind it and having undergone a rigorous peer review, cannot compensate for fluky data.
Additionally, due to the pressure on academics to publish papers in an effort to stay relevant in their respective fields – hence the adage ‘publish or perish’ – there are an unknown number of papers which have been published before replication experiments have been carried out.
All things considered, every retracted paper is a blow to science.
Image: John-Mark Kuznietsov