open source, open research, open publishing, SARS-CoV-2, COVID-19, long tail, curiosity, innovation

The global pandemic caused by the SARS-CoV-2 virus currently exceeds 8.5M cases and without reliable preventative treatments or a vaccine, the mortality rate from COVID-19 has already passed 450K. In response, and over the last six months an enormous volume of academic and clinical research has emerged online. PubMed^® is the largest search engine for biomedical literature and comprises more than 30M citations for literature from MEDLINE and other journals and publisher sites like PubMed Central^®.^1,2 PubMed’s role as the dominant resource for access to scholarly communication is widely acknowledged³ but faces increasing competition from open source alternatives.⁴ Research and development including the drive to publish is a labour-intensive activity requiring a high skill set, access to resources and time commitment in order to make scientific discoveries. Historically, the dominant and most familiar publishing model for the majority of traditional Journals listed on PubMed^® involves scientists using public or private funds to pay for the research, who in turn pay for publication, and then pay again to read the final published works.⁵ Preprint servers accelerate scholarly publishing through open source where content is not behind paywalls like traditional scientific publishers and feedback is encouraged prior to formal peer-review. The main benefit of this publishing pathway is immediacy and the ability to rapidly introduce new information into the academic community. Disadvantages are that search and retrieval can be more difficult, while a lack of or inadequate peer review could allow incorrect findings to be introduced or shared. These effects are likely reflected in the variable adoption pattern seen for some fields of study versus others.⁶ Nevertheless, scientific publishing in preprints⁷ and into lesser known Journals is increasing.⁸ In fact, the trend towards publishing as open access versus the traditional Journal pathway is based on four factors including visibility, cost, prestige and speed.⁹ In most cases, open source publishing offers undeniable cost-benefits and speed to publication.

In 2004 a concept called the Long Tail was popularised by the editor of Wired magazine.¹⁰ In this famous essay, the digital future was summarized as: “Forget squeezing millions from a few megahits at the top of the charts. The future of entertainment is in the millions of niche markets at the shallow end of the bitstream”. Through the lens of the long tail we can view PubMed® as the aggregator of ‘hits’ since they are most easily found and consumed. On PubMed ^®, the topic of the article as well as its publication date determine popularity.¹¹ But what about all the other research published elsewhere on preprint repositories – do these fulfil the criteria of ‘niches’ making up an ever-expanding long tail?

The purpose of this Letter is twofold. First, it is to highlight the power law scaling distribution for COVID-related publications available across different preprint servers at two time points (Figure 1). Second, it is to highlight how concepts like the long tail can be used to explore niche preprint repositories as a source of innovation that could lead to cross-disciplinary opportunities for novel discoveries. This second consideration is based on remix culture, where curiosity, knowledge sharing and reuse is a generator of new discovery.¹²

Figure 1 The main graph shows the keyword search for the term “COVID” performed over all the Paper Server.

Repositories listed in Table 1 spanning a 2-week period were then used to rank the number of COVID papers as the average. The long tail is evident towards the right-hand side of the distribution. Inset on log-log axes for this same data set, a power law emerges shown with an exponential least-squares fit having r² of 0.94 and power of -0.19. For comparison, a straight-line regression showed an r² of 0.90 and power of -2.84. Removing the PubMed^® data point caused the exponential fit to increase with an r² of 0.95 and a power of -0.18; similarly for the straight line fit, the r² was 0.91 and power -3.08 which preferentially captured the open source papers. Open circles and triangle symbols are used for the earlier and later dates respectively for the inset graph which also emphasizes that the niches are filling quickly.

The COVID crisis presents a new tipping point for scientific publishing according to Unesco¹³ and a recent analysis showed that COVID preprints are distributed at least 15-times more than non-COVID preprints.¹⁴ Currently, there is high-motivation for authors to generate novel research and produce public-good contributions, together with an expanding pool of servers ready to digitally disseminate this information (Table 1). Since scholarship and new idea formation is built on the work of others, when information costs reduce, the opportunity to accelerate new ideas and generate discoveries expands. It is my position that the different preprint servers act as digital aggregators which pre-filter content according to topic niche and act as post filters by assigning for example a Digital Object Identifier (DOI) or other digital tag that enables the object (the research paper) to be promoted on blogs and other social networks which is a type of recommender system. The repository is therefore the connector between supply and demand. Importantly, collaborative innovation within a Company¹⁵ (Pfizer) has been shown to follow a power law where rank frequency plots similar to Figure 1 for new ideas showed strong straight-line power laws with exponents spanning -2.7 to -3. This observation supports the claim that COVID research and publication output, regardless of where it resides on the internet emerges as a type of collective intelligence. At this point and with a solution-focused mindset across the range of potential problems surrounding the impact of COVID on society it is worth repeating the 8 key questions about curiosity and innovation that were developed by the Defense Advanced Research Projects Agency (DARPA) to assist with the discovery process.¹⁶ In practice, extracting and contributing new meaning into and out of the niches (or the hits) would take advantage of the 8 steps as a catalyst for novel idea generation and publication of same. These are: (1) What are you trying to do? Articulate your objectives using absolutely no jargon. (2) How is it done today, and what are the limits of current practice? (3) What is new in your approach and why do you think it will be successful? (4) Who cares? If you are successful, what difference will it make? (5) What are the risks? (6) How much will it cost? (7) How long will it take? (8) What are the mid-term and final “exams” to check for success?

I will conclude this letter by highlighting again from a recent article in Wired magazine reporting how virtual workspaces induced by the COVID-driven work from home is leading to not less but more research, and a greater degree of cross-disciplinary collaboration.¹⁷ Since niche papers are accessible over the internet at low to no cost to the reader, collaborative filtering on topic areas of interest will always ensure a small fan club. It is also very probable that scientists will increasingly need to exploit social media tactics and influencers to promote discoveries (or niche papers) to expand their audience and reach.¹⁸ The desire to contribute, innovate and communicate will always produce the usual ‘hit’ papers, but the long tail suggests that new and different ideas will emerge from the diversity and ambition written into the papers published in the ‘niches’.

None.

The author declares no conflicts of interest.

No funding.

1. PubMed. PubMed. 2020.
2. PubMed Central^®. Ncbi.nlm.nih.gov. 2020.
3. Williamson PO, Minter CIJ. Exploring PubMed as a reliable resource for scholarly communications services. J Med Libr Assoc. 2019;107(1):16–29.
4. Gasparyan AY, Yessirkepov M, Voronov AA, et al. Comprehensive Approach to Open Access Publishing: Platforms and Tools. J Korean Med Sci. 2019;34(27):e184.
5. Gierasch LM. On the costs of scientific publishing. J Biol Chem. 2017;292(39):16395–16396.
6. Chiarelli A, Johnson R, Richens E, Pinfield S. Accelerating scholarly communication The transformative role of preprints. 2019.
7. Matthew B Hoy. Rise of the Rxivs: How Preprint Servers are Changing the Publishing Process, Medical Reference Services Quarterly. 2020;39:84–89.
8. Bohannon J. Uprising: Less prestigious journals publishing greater share of high-impact papers. Science AAAS. 2020.
9. Conte S. Making the Choice: Open Access vs. Traditional Journals. 2020.
10. Anderson C. The Long Tail. WIRED. 2020.
11. Smith L, Wilbur W. The Popularity of Articles in PubMed. 5: 1-7.
12. Flath CM, Friesike S, Wirth M. et al. Copy, transform, combine: exploring the remix as a form of innovation. J Inf Technol. 2017;32:306–325.
13. Open access to facilitate research and information on COVID-19. UNESCO. 2020.
14. Fraser N, Brierley L, Dey G, et al. Preprinting a pandemic: the role of preprints in the COVID-19 pandemic. Biorxiv. 2020
15. Woods T. The Long Tail of Idea Generation. International Journal of Innovation Science. 2010;2(2):53–63.
16. Isaacs K, Ancona D. 3 Ways to Build a Culture of Collaborative Innovation. 2019.
17. Majumder M. Coronavirus Researchers Are Dismantling Science’s Ivory Tower—One Study at a Time. Wired. 2020.
18. Galetti M, Costa-Pereira R. Scientists need social media influencers. Science (New York, N.Y.). 2017;357(6354):880–881.