In transition

"Vent de llebeig, molta mar i peix fresc" (Always opportunity with crisis)

The time observer

By Albert Cardona, January 22nd, 2022.

Scientific research and its evaluation, entangled with scientific publishing, is in transition. We are living through a period of uncertainty where overall direction isn't yet set in stone and, therefore, opportunities abound for momentous changes.

My undergraduate lecturers in the late 1990's expressed great enthusiasm, even certainty, at the future of DNA vaccines--what today we have as mRNA vaccines for COVID-19. In those times, gene therapy was on the rage with few select success examples [1] fueling the collective imagination of those in the biomedical field. More than 20 years later, the seemingly overnight success of the mRNA COVID-19 vaccines builds on several decades of prior research that happened to reach maturity at the right time to save the day.

What we, the undergraduate students, were excited about, was genetic engineering. A promising technique of tantalizing possibilities, a colleague of mine and myself attended a seminar once that, we thought, would teach us the application of genetic engineering. Instead it taught us the dangers of its indiscriminate and reckless application, such as to crops and animals that, inevitably, interacted with wild populations and passed those traits on to them. In retrospect, genetic engineering has now become standard practice in the biology lab, both in industry and academia, but its application to the clinic are far and between. Turns out, most illnesses are multifactorial, which is a nice way to say that the "one gene, one protein, one disease" model has fallen very short of expectations.

Back then, as undergraduates in the late 1990's, we forecasted lab-grown foods. Having read lots of science fiction and also being enrolled in many biology subjects, our minds were primed. Obviously, plucking a muscle neuron, dedifferentiating it, expanding it in vitro, and triggering its multiplication and organization into muscle in vitro was feasible. No laws of physics against it, only our own lack of understanding. We fantasized, such Petri dish grown meats would self-wrap by growing a layer of edible, or cookable, plant tissue–like cardboard–on the outside, to self-package. Today we got instead bioreactor-grown bacteria or yeast that express cow protein and are later aggregated into "impossible burgers". A far cry from what we envisaged, but clearly on that trajectory. Why didn't we work on this field and became rich beyond imagining? The apparent obviousness and prospect of tedious, intellectually mediocre work necessary to deliver lab-grown foods were quite the deterrent to our younger selves.

More amusingly, after lectures on auxin and gibberellin hormones and plant development, the thought occurred to me that we could, someday, grow a house from a seed. A kind of tree whose shape would stabilize delineating walls and floors and ceilings and even windows via apoptosis, that is, programmed cell death. Why not? Yet, our understanding of development is limited, and such a living house organism is nothing but fantasy right now. In Richard Hamming's famous words ("You and your research"), we lack "a reasonable attack". Would address quite a few society-wide housing and related problems of 2022, including poor insulation, carbon capture, unavailability, and the increasing costs of labor. Alas, we lack an angle of attack onto the problem of designing the fast and shapely growth of a tree to become a house, so it will remain a pipe dream for the foreseeable future. Simply, we don't know anywhere near enough, despite the decades that, like sunrises and sunsets, come and go.

Science, as a body of knowledge, has become too large for anyone to comprehend. The last individual to know everything has long been dead, nearly two centuries. But as practitioners of science, we pretend we know enough, embracing a form of ignorance, or naivete, as the means to escape local maxima and initiate hill climbing towards new discoveries and great rewards (the "read too much" effect, by Hamming, again).

On knowing, what we surely don't know is how to evaluate and reward scientific research. We pretend the work reported in scientific papers has clear cut attributions to one or few authors. Bigger lies, or doctrine, or illusion, are hard to come by. Scientific research today is very much team work, and attribution is complicated to the point that it shouldn't even be attempted. Why would academia opt to continue to operate under XIX century terms may have to do with lack of change in how salaries are earned and careers advanced. The pressure to publish and take credit is higher than ever. When I turn down offers to join as author in papers for which I've barely done anything at all, I get puzzled looks. As if maximizing one's CV was the goal, rather than advancing our collective knowledge.

We biologists would do well to adopt the model of physicists and mathematicians. Publish as preprints and only, if needed at all, in journals, pretty much as an afterthought, if anything to provide journalists who can't evaluate scientific publications by themselves something that seems rubber-stamped as valid and valuable. And while at it, order authors alphabetically: no more chest-thumping noises on being first author or competing on who is to be first-first when sharing co-first authorship. As if this meant squat at all. In a faculty position call, when a search committee trims the list of applicants, such details matter not at all. When a candidate to a faculty position is interviewed, the good interviewer asks: "and what part did you do yourself personally for this paper?" Answers to this question make or break careers. Hand waving about the journal name disqualifies, or should disqualify, any applicant, under suspicion of conceited deceit. Consider that if the scientists who reported on the detection of gravitational waves with LIGO followed the biologists' approach to authorship, the disputes would still be ongoing and the binary black hole merger paper wouldn't have been published yet. The cost of not publishing groundbreaking work is astronomical: in delays, in broken careers, in opportunity costs of projects not started.

After all, preprints are freedom. Freedom from gatekeeping: no more desk rejects, no more attempting to reconcile disparate reviews, no more rebuttal letters to the editor, no more entombment of scientific research in the name of exclusivity for a journal. Preprints means Lynn Margulis wouldn't have had to wait for a decade and a half to publish her groundbreaking work on the theory of an endosymbiotic origin of eukaryotic organelles (Lynn Sagan, 1967, "On the origin of mitosing cells"). Margulis paper was rejected "fifteen or so" times prior to publication (Archibald 2014, cited in Gray 2017). Preprints also means people with guts can confront entire fields of research and win: witness the "Voodoo MRI" (Vul et al., 2008. Voodoo Correlations in Social Neuroscience, later appearing here with a watered down title) paper and its wake-up call, with transformational impact on the analysis of data from functional imaging studies using fMRI. (The dead salmon helped a lot, too. Even won an IgNobel! [2]) Whereas the dreaded long tail of "bad" papers in a preprint archive are entirely irrelevant and a tiny cost to pay for liberating scientific communications.

Preprints also mean the end of exclusivity: one entity, such as Review Commons, can publish reviews, while another reviewing entity can publish another set of reviews for the same paper. And there isn't any conflict. This is nothing else than normalizing what, at the moment, is called post-publication peer review. This should be the only form of review, for no other form of peer review, short of independent reproduction of the findings, makes any sense whatsoever.

Ending the static condition of published papers ("copy of record", a harmful concept) is of immense worth. Throw its unhelpful and cryptic (for invisible) "erratums" to the trash bin of history. We live in the times of digital documents, seamless multi-author version control, and DOIs. Hence, authors are free to revise, to update, to republish. All facilitated by preprint servers such as the bioRxiv, with helpful timestamped revision numbers. As the saying goes, we are living in the future, but the future is not evenly distributed. Helping those stuck in a pernicious past is quite the duty, and may have to merely await the pass of time.

The problem of scientific publishing is, at present, intimately linked to the problem of evaluating scientists themselves. Of career progression, promotions, salaries, prestige, prizes, recognition. The perverse incentives generated from such entanglement dominate modern scientific research. And addressing them requires courage, always in short supply.

Worse still, academic research is in a strange downward spiral. Governments want to fund discovery, but expect deliverables in 2-year time slots. Scientists that don't deliver under such absurd short term timelines get sidelined, as happened to Katalin Karikó who famously led to the development of mRNA vaccines for COVID-19. We have her tenacity and resourcefulness to thank, for what she endured is a scandal. Likewise for Emmanuelle Charpentier and the CRISPR targeted gene editing technology. When did academic scientific research go astray? I don't know, but Yuri Lazebnik's theory is intriguing: the impact of neoliberal doctrine on universities turned them into a chimeric monster between business and academia, and reduced academic scientists to "human resources", valued less than the very administrators that rule over them and whip them, as if creativity flourished under such conditions. With many reports that administrator positions are growing far faster than faculty positions [3], the corporate boat race joke sadly comes to mind.

We are living through interesting times. But aren't they all.

  1. Cystic fibrosis, a recessive gene disorder, comes to mind: "Correction of the ion transport defect in cystic fibrosis transgenic mice by gene therapy", Hyde et al. 1993.
  2. See also: "IgNobel Prize in Neuroscience: The dead salmon study".
  3. "Yale now has more administrators than undergrads thanks to a mammoth bureaucracy".