Publishing with eLife: “We chose eLife for an important philosophical reason…”

I am a late-career Professor of Immunology at Vanderbilt University, employed by the Medical Center corporation, after training as an MD-PhD physician-scientist and incidentally a clinician in Rheumatology. Since starting with post-doctoral training, my primary focus has been on identification of new basic molecular mechanisms and processes that influence the gene expression, survival, proliferation, and differentiation of lymphocytes. These interests led to a wave of work showing that external nutrient concentrations and intracellular metabolism can regulate lymphocyte differentiation and function, and to recognition of a glaring gap in bioscience – the inability to map metabolites in space. The eLife Reviewed Preprint is a step toward filling that gap.

The pragmatic reason is that funding for this project was essentially at an end. The referee mindset of demanding more and more has spread to journals at almost all levels of perceived importance…

A personal peeve or philosophical bias is that biochemical entities need to be approximated or (semi)-quantitated to improve and advance the quality of research on particular topics. Of course, nothing is perfect, and sometimes what is needed is to push the technological frontier. There now is very compelling, even overwhelming, evidence of how nutrients and intermediary metabolism can regulate cellular physiology, and indications of localized effects for cells in their micro-environments in vivo.

However, the distributions of such molecules within tissues has been linked to function in few, if any, instances, and a key technical limitation is that the distributions cannot be measured. So at the moment the literature is littered with hand-waving. For technical reasons, (phospho)lipids are more readily mapped by imaging mass spectrometry performed with tissue sections. So that was a point of departure for the work we present here.

In an initial study (Analytical Chemistry, 2020), to show how the technology can be used in complex tissues with a wide variety of distinct cell types packed closely together in space, we had performed lipidomics with spleen sections in which a unique micro-anatomical structure involved in immunity and vaccine efficacy - the germinal center (GC) - was identified by use of a transgenic mouse line that encoded a fluorophore whose expression is driven to high levels uniquely in B cells.

Surprisingly, the work identified enrichment of a type of phospholipid that is rarely thought about outside a rather small silo, the ether phospholipid. The present study was designed to test the hypothesis generated by this initial (2020) lipidomic study. To do so, we were fortunate to be able to collaborate with and draw on ground-breaking work of C. Semenkovich and colleagues, who had previously generated and published a paper using a conditional loss-of-function (“knockout”) allele in a different lineage of white blood cells.

We sought to answer a few simple questions on which almost no information is available.

(1) Would depletion of an enzyme important in biosynthesis of ether lipids affect the relative amounts of phospholipids in GC, lymphoid follicles, or ex vivo if restricted to one cell type (the B cell)?

(2) Would there be evidence that such biosynthesis in situ affected B cell physiology, GC, or antibody responses?

We feel that answering these questions in the affirmative would provide strong evidence that spatial imaging of metabolite concentrations could identify new processes or druggable targets (in this case, the enzyme PexRAP, also known as ADHAPR, acyl/alkyl DHAP reductase).

Philosophically … the journal, editors, and referees at least know that their comments will be publicly viewable AND – importantly – authors can comment and, if needed, rebut or refute false claims

We chose eLife for an important philosophical reason (touched on in the News in Focus section of Nature, Jan 9, 637: 258-259) and a rather pragmatic one. A simple way of putting it is that we liked the journal’s broad scope, reasonable reputation, and model for peer review and dissemination of insights after rounds of quality filtering. The pragmatic reason is that funding for this project was essentially at an end. The referee mind-set of demanding more and more has spread to journals at almost all levels of perceived importance, and, worse yet, there too often is no real accountability of referees or editors (especially not at what Bob Roeder once wittily termed “fashion magazines”). Because the manuscript and findings are not “just immunology” and instead are aimed at a broader audience and general question, we wanted a “general biological sciences” journal but also one where some predictability would apply. The topic editor would be a true expert (in this case, in work on B cells or GC). If they felt the manuscript insufficiently interesting or obviously flawed on the face of it, it would not go out for review, and we could move on.*

*(Editor’s note: How and why eLife selects papers for review is explained here.)

Philosophically, a feature of the new eLife Model and its predecessor that is attractive is that the journal, editors, and referees at least know that their comments will be publicly viewable AND – importantly – authors can comment and, if needed, rebut or refute false claims that too often appear in reviews at high-profile or general-interest journals where the reviews never see the light of day. Inherent shortcomings of peer review are exacerbated by (among many factors) the lack of accountability combined with insufficient time (excess haste), reputational competition, and “technical ‘virtue-signaling’”. (For a while, almost nothing could suffice unless it had scRNA-seq or a wide-net CRISPR screen. Now, perhaps that is too common and only with spatial or spatial scRNA-seq and some multi-omics or ‘systems immunology’.)

A consequence of undue pressures linked to journal “brand” and reputation – or worse yet “citation impact factor” of the journal – is that taxpayers who fund the bioscience enterprise are stuck with a system that impedes the open flow of information on results that support a novel conclusion or insight but lack the bells and whistles, or are delayed by the fact that “more always could be done”. The new eLife Model offers something a step better than the “preprint server” model (e.g. bioRxiv) in that a top scientist will have vetted the work, and then the public “preprint” gives readers takes on how the manuscript could be better. I believe that support of this new model from the scientific community is important.

I really like the transparency. Funding agencies should use the power of the purse to push ALL peer-reviewed journals to post the referee reports at each cycle of review…

At present, we are at the stage of a publicly Reviewed Preprint (with authors’ provisional response). We have elected to try to get a bit more work done and to do some text editing to draw on the comments of the two referees.

Main impression – so far, so good. The logistics of the process were similar to getting started with traditional publishing. The process went fairly smoothly. The turn-around was a bit slower than what a terrific “specialty journal of record” – the J Immunol – achieves or than with “fashion magazines” (which ask for referee reports within, say, 10 days for submissions with 100 figure panels and dense, complex work). But a personal view is that so much slop (referee errors or oversights) makes it through the over-rushed process as to make it preferable to have a slower turnaround but a predictable path.

A big positive with eLife turned out to be prompt and helpful responses from journal staff answering questions from a “newbie”, providing an extra week for authors to send in the provisional response and clarity about the issue of “final article of record”.

Having learned the name of the editor who made the decision to send our work out for review, and respecting them immensely, I would say that at least the send / no review decision is not terribly different from fashion magazines (“CNS journals” and the most prominent of their spawn, i.e., the name brand spin-offs) but is more transparently a decision of a major expert in the topic.

More broadly, I really like the transparency. Funding agencies should use the power of the purse to push ALL peer-reviewed journals to post the referee reports at each cycle of review (1st submission; revision), and accompany those reports with the authors’ public reply (perhaps edited or moderated in case of inappropriate personalization or heat). Ideally, referees would include what they consider the limitations – while being mindful of what the authors actually were setting out as the advance of the paper.

Another of the major challenges for science – in this era where the Endarkenment encroaches on us – is to do better at (re-)building trust, for which transparent mechanisms of post-publication peer review are essential (and ought to be mandatory for journals that receive page charges derived from taxpayer-funded research grants).

Since we will submit a revision and in principle enlist the same referees, answering this question is not without some wrinkles (challenges).

On the whole, the initial review was mostly lucid in that the editors made matters fairly clear to us. I thought the eLife Assessments were at least as good as or better than the median of review quality experienced with well-regarded journals in the immunology silo. The good news is that compared to some journals that see themselves as benchmarks, yet the referees and editor make outright errors or distort facts, the eLife reviews were appropriate and mostly revolved around a subjective question of “when is enough enough?” and a cognitive filter of focusing on the manuscript as exclusively reporting about GC physiology. One report was rather short, light on details, and a bit ambiguous (though of course I would like to think that is because on the whole the reviewer liked the body of evidence and its presentation). The other review focused greatly on the germinal center aspects of the work, a cognitive processing that is par for the course, but on the whole the review was rather like that of a high-profile journal.

...most of my papers have been enhanced by peer review input, but sometimes that is not the case and the eLife Model gives authors more latitude to decide...

I encourage giving it a try with some of your better or best work if it does not get reviewed (or gets reviewed and dissed) at the conventional “tip-top” journals (CNS and the most prominent of their spin-offs). I especially suggest doing so instead of paying an almost insane premium to publish at vastly higher cost in the lower-level spin-off journals of the CNS ecosystem that are classics of the business (a word properly applied here) of “name branding”. I am aware of multiple papers that – in my opinion – the editors at high-profile, name-brand journals screwed up on, in either not sending out for review (from not understanding the advance) or in paying undue attention to spurious comments of a reviewer. There is a kernel of accuracy in the complaint of some science advocates and members of the taxpaying public about paying over two-fold more in a process that inherently delays the dissemination of important insights.

In addition, an esteemed colleague commented that once reviewers get hold of a paper, one loses control, or the focus, of a story. Of course, most of my papers have been enhanced by peer review input, but sometimes that is not the case and the eLife Model gives authors more latitude to decide, readers more information about the work, and the public who fund our work more insight into the give and take of scientific research.

Quite a lot needs to change. There is nothing like the wave of demagoguery and the current convulsions here in the USA to illustrate dire threats because of shortcomings in the bonding of scientists and of science to the public whose understanding and support (including but not limited to funding) are preconditions for what we get to do.

the power dynamic and current beneficiaries mean that we likely will stay stuck with a flawed system in which a large amount of the prestige ecosystem is dominated by for-profit publishing companies

Having said that, where matters stand with research publication and valuation is rather like the medical-industrial complex in the USA (inaccurately referred to as “health care system”) – a majority feels and knows it needs to change, but the power dynamic and current beneficiaries mean that we likely will stay stuck with a flawed system in which a large amount of the prestige ecosystem is dominated by for-profit publishing companies. To recap issues I touched on above:

a. First rectify the language. Authors should uniformly shun, and journals should ban, titles, abstracts, and writing in general that make it seem as though one story, one bundle of data yielding the authors’ conclusions has defined a ‘universal truth’, a generally applicable advance or understanding. This mindset of authors, especially many accustomed to regular publication in high-profile or trend-dictating journals, contributed very substantially to current problems in ways that started during the pandemic and involved an unhealthy intersection of “popular press” / conventional media, ego, and scientists.

b. If eLife, for $2,500, can afford the server / cloud space to post and make fully accessible the referee reports at each cycle, and – ideally – referee comments on the remaining shortcomings in or limitations of an accepted final version of an article, so can Springer-Verlag (Nature brand; NPG), Elsevier (Cell brand), and AAAS.

c. A lot would need to change in the whole culture of scientists. Charlie Janeway, during my postdoctoral fellowship, tried to have a journal with “radical transparency” – I think names of referees as well as the reviews. My postdoc supervisor accurately noted that authors and referees would shun it.

d. More broadly, the over-hasty time limits imposed on referees, and the inherent limits of any two or three referees in the pre-publication peer-review model point to the need for a better and curated mechanism for POST-publication peer review (again, more or less imposed on journals and authors by the funding agencies that, after all, pay most of the page charges AND a lot of the subscription costs…). Curation would be needed so as to have a way of keeping it to PEER review, avoid the inevitable deluge of trolls, prevent grinding of axes, etc.

Best of luck with the model. Changing a culture is very hard. This fact is particularly true when the changes might gore the big oxen of reputation-based assessments. To adapt from one wag's line, “evidence-based decisions” ought to be the practice rather than “eminence-based decisions”.

“This model has many benefits in my mind. But one of the biggest advantages is that the authors have a lot of control in the publication process.” – Meike van der Heijden

“We liked the idea of having an open "conversation" with the reviewers...without the threat of rejection.” – Patrick Allard

“... the review and publication process with eLife was smooth, fair, and transparent.” – Ushio Masayuki

How publishing with eLife works
What is a Reviewed Preprint?
What is an eLife Assessment?
What happens after you submit your research?