It’s another beautiful day on the upper east side of Manhattan. The sun shines through the window shades, my 2-year-old daughter sings to herself as she wakes up, my wife has just returned from an early-morning workout – all is right as rain.
My phone buzzes. My stomach clenches. It buzzes again. My Twitter alerts are here. I dread this part of my morning ritual – finding out if I’ve been scooped overnight by the massive inflow of scientific manuscripts reported to me by my army of scientific literature–searching Twitter bots.
That’s right, Twitter isn’t just for presidents anymore, and in fact, the medical community has embraced Twitter across countless fields and disciplines. Scientific conferences now have their specific hashtags, so those of you who couldn’t come can follow along at home.
But this massive data dump now has a #fakenews problem. It’s not Russian election meddling, it’s open source “preprint” publications. Nearly half of my morning list of Twitter alerts now are sourced from the latest uploads to bioRxiv. BioRxiv is an online site run by scientists at Cold Spring Harbor Laboratory and is composed of posting manuscripts without undergoing a peer-review process. Now, most commonly, these manuscripts are concurrently under review in the bona fide peer-review process elsewhere, but unrevised, they are uploaded directly for public consumption.
There was one recent tweet that highlighted some interesting logistical considerations for bioRxiv manuscripts in the peer-review process. The tweet from an unnamed laboratory complains that a peer reviewer is displeased with the authors citing their own bioRxiv paper, while the tweeter contends that all referenced information, online or otherwise, must be cited. Moreover, the reviewer brings up an accusation of self-plagiarism as the submitted manuscript is identical to the one on bioRxiv. While the latter just seems like a misunderstanding of the bioRxiv platform, the former is a really interesting question of whether bioRxiv represents data that can/should be referenced.
Proponents of the platform are excited that data is accessible sooner, that one’s latest and greatest scientific finding can be “scoop proof” by getting it online and marking one’s territory. Naysayers contend that, without peer review, the work cannot truly be part of the scientific literature and should be taken with great caution.
There is undoubtedly danger. Online media sources Gizmodo and the Motley Fool both reported that a January 2018 bioRxiv preprint resulted in a nearly 20% drop in stock prices of CRISPR biotechnology firms Editas Medicine and Intellia Therapeutics. The manuscript warned of the potential immunogenicity of CRISPR, suggesting that preexisting antibodies might limit its clinical application. Far more cynically, this highlights how a stock price could theoretically be artificially manipulated through preprint data.
The preprint is an open market response to the long, arduous process that peer review has become, but undoubtedly, peer review is an essential part of how we maintain transparency and accountability in science and medicine. It remains to be seen exactly how journal editors intend to use bioRxiv submissions in the appraisal of “novelty.”
How will the scientific community vet and referee the works, and will the title and conclusions of a scientifically flawed work permeate misleading information into the field and lay public? Would you let it influence your research or clinical practice? We will be finding out one tweet at a time.
Aaron D. Viny, MD, is with the Memorial Sloan Kettering Cancer Center, N.Y., where he is a clinical instructor, is on the staff of the leukemia service, and is a clinical researcher in the Ross Levine Lab. He reported having no relevant financial disclosures. Contact him on Twitter @TheDoctorIsVin.
Photo by Gina Motisi, Cold Spring Harbor Laboratory