US research censure stirs up Twitter

Journal name:
Date published:
Published online

Some compelling papers have been through the social-media mill — and not all of them survived.

Based on data from Based on data from Altmetric is supported by Macmillan Science and Education, which owns Nature Publishing Group.

The US biomedical-research system is overcrowded, underfunded and headed for decline, according to a much-discussed report by four top-tier researchers, including Nobel prizewinner Harold Varmus, director of the US National Cancer Institute. The authors argue that the imbalance between the surge of talented scientists and the steady dwindling of funds for research has created a “hypercompetitive” atmosphere that stifles careers and hinders scientific progress. The paper resonated with young researchers, guaranteeing significant coverage on Twitter. John Bachman, a graduate student in a systems-biology lab at Harvard University, tweeted that the paper was a “must-read”. “Young researchers see different parts of the situation,” Bachman said in a phone interview. “This paper connected the dots between policy and the academic job market. It really pulled the picture together.”

Alberts, B., Kirschner, M. W., Tilghman, S. & Varmus, H. Proc. Natl Acad. Sci. USA 111, 57735777 (2014)

Can social-media chatter affect a paper's legacy? In a timely piece, biologist and blogger Zen Faulkes of the University of Texas Pan-American in Edinburg looks at peer review in the age of Twitter. As he notes, fellow scientists now have a chance to publicly praise or disparage papers long after publication. Social media, he says, is especially adept at highlighting studies with serious flaws, including the now-retracted paper by Gilles-Éric Séralini et al. that purported to find links between genetically modified maize (corn) and cancer in rats (G. E. Séralini et al. Food Chem. Toxicol. 50, 42214231; 2012). To make the circle complete, Faulkes's paper is now making the rounds on Twitter. Canadian physician Todd Raine singled out an especially memorable quote: “Postpublication peer review on social media is like the mosh pit at a punk rock show.”

Faulkes, Z. Neuron 82, 258260 (2014)

Some papers get a new lease of life on social media years later. Just look at the buzz that still surrounds Stanford University researcher John Ioannidis's classic 2005 paper, 'Why Most Published Research Findings are False.' The paper used research simulations to conclude that “for most study designs and settings, it is more likely for a research claim to be false than true”. Ioannidis argues that small sample sizes, small reported effects and less-than-rigorous designs all increase the odds that a so-called finding doesn't represent reality. PLoS Medicine, where the paper was published, recently tweeted that the paper has surpassed more than 1 million views, putting it in the limelight once again. The article's resurgence on social media prompted Muin Khoury, director of the Office of Public Health Genomics at the US Centers for Disease Control and Prevention, to tweet an observation that is as true today as it was in 2005: “A little replication goes a long way.”

Ioannidis, J. PLoS Med. 2, e124 (2005)

Not many papers end up getting compared with an April Fool's joke, but a recent uncorrected proof at Drug Discovery Today had bloggers scratching their heads and grasping for explanations. Calvin Yu-Chian Chen of China Medical University in Taiwan made the eye-catching claim that it could soon be possible to “print any drug instantly” on a 3D printer. Chen envisions a “DrugPrinter” reaction chamber based on a “traditional Chinese egg-cake oven” that could construct drugs at the atomic level. He writes that he could develop a prototype within five years, but the paper is scant on details. Echoing a theme among several science blogs, In The Pipeline called the paper “a fantasy of atomic-level nanotechnology” and “one of the oddest papers to appear” in Drug Discovery Today. Chen joined the discussion on In The Pipeline and other blogs, repeatedly saying that he asked the journal editor to withdraw the paper — not because he disavows the claims, but because he wants to stop the ridicule he is getting. In an e-mail, Chen speculated that this might be the first time that a paper was withdrawn solely because of blog posts. “This idea is indeed crazy, however, most of the tech is feasible right now,” he says. “I just can't stand the attitude in the blogs.” Chen got his wish — the journal has since withdrawn the paper at his request.


Sometimes, blog posts can help to save interesting papers from obscurity. When Discover magazine blogged about a new computer algorithm that is better than humans at recognizing faces, social media took notice. The arXiv paper by computer scientists Chaochao Lu and Xiaoou Tang at the Chinese University of Hong Kong was practically guaranteed decent buzz after @Neuroskeptic (also a Discover blogger) tweeted the blog post to his 36,000 followers. Whereas some on Twitter hailed the technological achievement, @billfranksga had a different take: “#bigbrother here we come!”

Lu, C. & Tang, X. Preprint at (2014)

Tao Ding, a postdoc at the University of Michigan, waited for a prime occasion to post his first ever tweet: publication of his Nature Letter outlining the interplay between a person's background and his or her microbial community. Ding and his colleague Patrick Schloss found that certain types of community largely depend on a person's gender and whether they were breastfed as infants. Intriguingly, the researchers also found that people with different levels of schooling tended to have different microbiomes, an association that definitely got noticed by Twitter users — some of whom undoubtedly wondered which germs they picked up at university.

Ding, T. & Schloss, P. D. Nature (2014)

Additional data