The Looming Problem of Word Limits and Scientific Publication

December 7, 2009 · Posted in HUGO President Blogs · 1 Comment 

…or Length Matters

by HUGO President Prof. Edison T Liu

I have been confronted both as an author and a reviewer with the difficulty of explaining a complex story within a 1500 or even a 5000 word limit for a manuscript.  The basis of the word limits in many journals have been historical , and often was because of print costs.  But as science advances, this publication boundary has not moved with the times.  What exactly have been the trends in the biological sciences? 

Genomics has been the vanguard of these trends.  The datasets are massive, the analyses are complex, and the validation is also extensive.  When a high throughput screen for, let’s say, siRNA knock downs, or for synthetic lethality in a genome-wide scale is presented, the appropriate description for just the initial set up and analysis may run into the maximum 5000 words.  The biological interrogations and functional validations can rationally require many more words to explain.

In order to compress the presentation, authors have rendered figures with many subpanels and place critical data in supplementary sections that can be many times the size of the original manuscript.  In one systems biology paper (Pujana MA, et al. Nat Genet. 39(11):1338-49, 2007), one figure (figure 4) had 4 parts (a-d) but presented 9 panels of high analytical complexity.  Our paper recently published in Nature (Fullwood MJ, et al. Nature. 462(7269):58-64, 2009), had a supplement with ~10,000 words for a manuscript that contained less than 5,000 words.  The HUGO Pan-Asian SNP Consortium will publish in Science (December 2009) a scientific story in 4000 words which will require a supplementary section of 14,000 words. 

It can be said that the supplemental section should be the place to explain the details not necessary for the core story.  However, this is becoming less true, especially in genomics, where the massive nature of the data and the analysis requires compartmented analysis each of which are critical to the main story.  Operationally, this is leading to a troubling trend.  The supplemental data is given marginal analysis by both readers and reviewers.   In many ways, this is like presenting each classical novel as a CliffsNotes summary with a supplement that contains the original chapters.  One possible outcome is that a distracted public and even expert review panels will miss serious problems in the analysis that is “tucked” away in the supplemental section (as implied in Coombes KR, Wang J & Baggerly KA, Nature Medicine 13, 1276 – 1277, 2007).

Indeed, our schedules are busy and our time is precious, but perhaps we need a publication format that will preserve the needed scholarship and yet respect the limited attention span of busy scientists.  For example, why not have an extended abstract of ~400-500 words that tells the core story (even include a figure or two), but allow for the detailed work to be presented in its entirety as a coherent narrative.  Given the online nature of journals, this surely cannot be more expensive than the current main-text-with-online-supplement format.  

Regardless of whether you are pro or con to this opinion piece, I would encourage the readers to comment and to engage their communities in such a discussion about change in scientific publishing.