PLOS Biology

Amidst the COVID-19 pandemic, preprints in the biomedical sciences are being posted and accessed at unprecedented rates, drawing widespread attention from the general public, press and policymakers for the first time. This phenomenon has sharpened longstanding questions about the reliability of information shared prior to journal peer review. Does the information shared in preprints typically withstand the scrutiny of peer review, or are conclusions likely to change in the version of record? We assessed preprints from bioRxiv and medRxiv that had been posted and subsequently published in a journal through 30th April 2020, representing the initial phase of the pandemic response. We utilised a combination of automatic and manual annotations to quantify how an article changed between the preprinted and published version. We found that the total number of figure panels and tables changed little between preprint and published articles. Moreover, the conclusions of 7.2% of non-COVID-19-related and 17.2% of COVID-19-related abstracts undergo a discrete change by the time of publication, but the majority of these changes do not qualitatively change the conclusions of the paper.

Editorial decision-making is a fundamental element of the scientific enterprise. We examined whether contributions to editorial decisions at various stages of the publication process is subject to gender disparity, based on analytics collected by the biomedical researcher-led journal eLife. Despite efforts to increase women representation, the board of reviewing editors (BRE) was men-dominant (69%). Moreover, authors suggested more men from the BRE pool, even after correcting for mens numerical over-representation. Although women editors were proportionally involved in the initial editorial process, they were under-engaged in editorial activities involving reviewers and authors. Additionally, converging evidence showed gender homophily in manuscripts assignment, such that men Senior Editors over-engaged men Reviewing Editors. This tendency was stronger in more gender-balanced scientific disciplines. Together, our findings confirm that gender disparities exist along the editorial process and suggest that merely increasing the proportion of women might not be sufficient to eliminate this bias.

The academic community has been increasingly using preprints to disseminate their latest research findings quickly and openly. This early and open access of non-peer reviewed research warrants new means from the scientific community to efficiently assess and provide feedback to preprints. Yet, most peer review of scientific studies performed today are still managed by journals, each having their own peer review policy and transparency. However, approaches to uncouple the peer review process from journal publication are emerging. Additionally, formal education of early career researchers (ECRs) in peer reviewing is rarely available, hampering the quality of peer review feedback. Here, we introduce the Preprint Club, a cross-institutional, community-based approach to peer reviewing, founded by ECRs from the University of Oxford, Karolinska Institutet and Icahn School of Medicine at Mount Sinai. Over the past two years and using the collaborative setting of the Preprint Club, we have been discussing, assessing, and providing feedback on recent preprints in the field of immunology. In this article, we provide a blueprint of the Preprint Club basic structure, demonstrate its effectiveness, and detail the lessons we learned on its impact on peer review training and preprint authors perception.

Despite longstanding recognition of sex as a biological variable, its integration into biomedical research remains inconsistent. Numerous publishers have introduced policies to improve reporting and inclusion of sex and gender, including Nature, which requires authors to complete a Life Science Reporting Summary documenting sex inclusion. Here, we evaluated the effectiveness of these policies by examining sex inclusion and reporting practices in all original research articles involving humans, vertebrates, or cell lines published in Nature in 2025 (N=513). Nearly two-thirds of articles included both sexes (62.7%); however, inclusion was often nominal. Of these articles reporting inclusion of both sexes, 33% did not maintain inclusion across experiments, used markedly unbalanced sex ratios ([≥]2:1), or alternated between male- and female-only experiments. Another 45.5% of these articles reporting inclusion of both sexes did not report sample size by sex, so it cannot be ascertained whether sex inclusion was maintained across experiments or balanced by sex. Single-sex studies accounted for approximately one-fifth of articles. While male-only and female-only studies occurred at similar overall rates, male-only studies were more than four times more likely to address conditions affecting both sexes while female-only studies were more likely to address sex-specific conditions (e.g., ovarian cancer). Only 7% of articles explicitly analyzed sex as a discovery variable for at least some analyses. These findings suggest that transparency-focused reporting summaries alone are insufficient to ensure sex inclusion and/or meaningful analytical integration of sex. As a leading biomedical journal, Nature plays a central role in shaping research norms; without stronger editorial expectations, reporting requirements risk reinforcing male-default assumptions rather than advancing rigor and generalizability.

This article presents the results of a survey of the three major multidisciplinary journals (Nature, Proceedings of the National Academy of Science, Science). Fifty articles involving experiments necessitating the agreement of an ethics committee were searched in each of the three journals to June 30th 2023. Because PNAs announced that a Statistical Review Committee was created and working since September 2023, a further set of 50 articles was retrieved and analyzed from 2024 January 1st. The following items were checked in articles: an explicit statement of approval from an animal ethics committee, the calculation of the appropriate samples size needed and how randomization and blinding were performed, the minimum sample size reported, the presence of repeats, and the methods used to limit type I and type II errors. No clear experimental design was fully reported in any article. The major problems were 1) extremely small sample sizes (<4/group) in nearly half of the articles, 2) confusion between biological repeats and technical replications, and 3) lack of correction for multiple comparisons. These errors led to major inflation of type I and type II errors. In conclusion, only 10 percent of the articles analyzed presented correct statistical methodology.

The open science movement aims to transform the research landscape by promoting research transparency in order to enable reproducibility and replicability, lower the barriers for collaboration, and reduce unnecessary duplication. Recently, in recognition of the value of open science, funding agencies have begun to mandate open science policies as a condition in grantee awards. However, operationalization and implementation of an open science policy can have unanticipated costs and logistical barriers, which can impact both the funder, as well as the grantee. These factors should be considered when implementing an open science policy. The Aligning Science Across Parkinsons (ASAP) initiative utilizes a comprehensive open science policy, which, in addition to requiring immediate free online access to all publications, also requires all newly-generated datasets, protocols, code, and key lab materials be shared by the time of publication. Moreover, preprints must be posted to a preprint repository by the time of manuscript submission to a journal for review. Here, we outline the potential costs associated with implementing and enforcing this open science policy. We recommend that funders take these considerations into account when investing in open science policies within the biomedical research ecosystem.

Withdrawal StatementThe authors have withdrawn this manuscript because additional authors were not listed and have not consented to the production of this preprint. Therefore, the authors do not wish this work to be cited as reference for the project. If you have any questions, please contact the corresponding author.

In a growing digital landscape, enhancing the discoverability and resonance of scientific articles is essential. Here, we offer ten recommendations to amplify the discoverability of studies in scientific databases. Particularly, we argue that the strategic use and placement of key terms in the title, abstract, and keyword sections can boost indexing and appeal. By surveying 237 journals in ecology and evolutionary biology, we found that current author guidelines may unintentionally limit article discoverability. Our survey of 5842 studies revealed that authors frequently exhaust abstract word limits -- particularly those capped under 250 words. This suggests that current guidelines may be overly restrictive and not optimised to increase the dissemination and discoverability of digital publications. Additionally, 91.9% of studies used redundant keywords in the title or abstract, undermining optimal indexing in databases. We encourage adopting structured abstracts to maximise the incorporation of key terms in titles, abstracts, and keywords. In addition, we encourage the relaxation of abstract and keyword limitations in journals with strict guidelines, and the inclusion of multilingual abstracts to broaden global accessibility. These evidence-based recommendations to editors are designed to improve article engagement and facilitate evidence synthesis, thereby aligning scientific publishing with the modern needs of academic research.

We conducted an international cross-sectional survey of biomedical researchers perspectives on the reproducibility of research. This study builds on a widely cited 2016 survey on reproducibility, and provides a biomedical-specific and contemporary perspective on reproducibility. To sample the community, we randomly selected 400 journals indexed in MEDLINE, from which we extracted the author names and e-mails from all articles published between October 1, 2020 and October 1, 2021. We invited participants to complete an anonymous online survey which collected basic demographic information, perceptions about a reproducibility crisis, perceived causes of irreproducibility of research results, experience conducting replication studies, and knowledge of funding and training for research on reproducibility. A total of 1924 participants accessed our survey, of which 1630 provided useable responses (response rate 7% of 23,234). Key findings include that 72% of participants agreed there was a reproducibility crisis in biomedicine, with 27% of participants indicating the crisis was significant. The leading perceived cause of irreproducibility was a pressure to publish with 62% of participants indicating it always or very often contributes. About half of the participants (54%) had run a replication of their own previously published study while slightly more (57%) had run a replication of another researchers study. Just 16% of participants indicated their institution had established procedures to enhance the reproducibility of biomedical research; and 67% felt their institution valued new research over replication studies. Participants also reported few opportunities to obtain funding to attempt to reproduce a study and 83% perceived it would be harder to do so than to get funding to do a novel study. Our results may be used to guide training and interventions to improve research reproducibility and to monitor rates of reproducibility over time. The findings are also relevant to policy makers and academic leadership looking to create incentives and research cultures that support reproducibility and value research quality.

Neurosecretory brain centers exist as part of the central nervous system of many bilaterian animals and play pivotal roles in the control of feeding and metabolism. In the animal evolution, the nutrition regulation appears to have been essential for the early heterotrophic metazoan ancestor, but the underlying evolutionary processes remain unclear. We found that the cnidarian Nematostella vectensis develops an oral/pharyngeal nervous system that shares a core signature with bilaterian neurosecretory brain centers comprising Orthopedia and neuropeptide RFamide-positive neurons, and that is essential for feeding regulation. Our data suggest that prior to the bilaterian evolution, the Orthopedia/RFamide-positive neural assembly was already functioning as a prototype of the neurosecretory brain center for the feeding regulation that could be the driving force of neural centralization.

This study investigated how physical luminance and perceived brightness affect breakthrough time (BT) under continuous flash suppression (CFS). Experiment 1 examined whether the glare illusion--which increases subjective brightness without altering actual luminance--would shorten BT compared to physically identical controls. The results revealed no difference, suggesting that subjective brightness alone does not expedite emergence into awareness. Experiment 2 assessed whether the partial suppression of the illusions inducers influenced detection speed and revealed that subjective brightness stimuli gained a BT advantage. Experiment 3 tested participants ability to discriminate real versus illusory brightness while stimuli remained suppressed; performance above chance for both conditions indicated that physical and perceived brightness cues were processed unconsciously. Together, these findings suggest that contextual brightness illusions are not simply lost below awareness--they can be discriminated in unconscious vision. Statement of RelevanceUnderstanding how the visual system processes brightness illusions--even prior to awareness--is crucial for uncovering the brains deeper mechanisms of perception. This research clarifies the limits of unconscious visual processing by showing that illusions can be registered without consciousness but do not necessarily expedite emergence into awareness. These insights advance theoretical models of hierarchical vision and demonstrate that while early cortical areas prioritize actual luminance in determining whether a stimulus reaches awareness, higher areas can still encode illusory brightness beneath the threshold of consciousness. Thus, the studys findings have broad implications for both basic neuroscience--refining how we think about the boundary between unconscious and conscious perception--and for applied fields seeking to harness or mitigate perceptual illusions. Research Transparency StatementO_ST_ABSPreregistrationC_ST_ABSAll experiments were not preregistered. Conflicts of interestThe authors declare that they have no competing interests. Data availabilityThe code and data underlying the results presented in the study are available from the Open Science Framework repository (https://osf.io/u5der/). EthicsThe Committee for Human Research of the Toyohashi University of Technology approved the experiments (2023-20). Written informed consent was obtained from all participants after the procedural details were explained to them. FundingThis work was supported by JSPS KAKENHI (Grant Numbers JP22K17987 to H.T., JP24H01551 to T.M., JP23KK0183 to T.M., JP20H05956 to S.N.), JSPS Grant-in-Aid for JSPS Fellows (Grant Number JP24KJ1313), and Young Principal Investigator fund JPMJFS2121. Artificial intelligenceDuring the preparation of this work, the authors used ChatGPT o1 and Grammarly to improve the language, and the manuscript was proofread by native English speakers through an English editing service. After using the tool and service, the authors reviewed and edited the content as needed, and they take full responsibility for the content of the publication.

The reality that volumes of published research are not reproducible has been increasingly recognised in recent years, notably in biomedical science. In many fields, spurious results are common, reducing trustworthiness of reported results. While this increases research waste, a common response is that science is ultimately self-correcting, and trustworthy science will eventually triumph. While this is likely true from a philosophy of science perspective, it does not yield information on how much effort is required to nullify suspect findings, nor factors that shape how quickly science may be correcting in the publish-or-perish environment scientists operate. There is also a paucity of information on how perverse incentives of the publishing ecosystem, which reward novel positive findings over null results, shaping the ability of published science to self-correct. Precisely what factors shape self-correction of science remain obscure, limiting our ability to mitigate harms. This modelling study illuminates these questions, introducing a simple model to capture dynamics of the publication ecosystem, exploring factors influencing research waste, trustworthiness, corrective effort, and time to correction. Results from this work indicate that research waste and corrective effort are highly dependent on field-specific false positive rates and the time delay before corrective results to spurious findings are propagated. The model also suggests conditions under which biomedical science is self-correcting, and those under which publication of correctives alone cannot stem the propagation of untrustworthy results. Finally, this work models a variety of potential mitigation strategies, including researcher and publication driven interventions. Significance statementIn biomedical science, there is increasing recognition that many results fail to replicate, impeding both scientific advances and trust in science. While science is self-correcting over long time-scales, there has been little work done on the factors that shape time to correction, the scale of corrective efforts, and the research waste generated in these endeavours. Similarly, there has been little work done on quantifying factors that might reduce negative impacts of spurious science. This work takes a modeling approach to illuminate these questions, uncovering new strategies for mitigating the impact of untrustworthy research.

Detecting causal language in scientific literature is critical for understanding how research fields frame evidence and inform interventions and policies, yet existing approaches commonly rely on manual annotation. The objective of this study was to evaluate four classifiers for detecting causal language and to apply the best-performing model to assess trends in microbiome research. Microbiome research, with its rapidly expanding observational literature, provides a relevant case study. We extracted Term Frequency-Inverse Document Frequency (TF-IDF) features from the last three sentences of available publication abstracts and trained four classifiers (L1- and L2-regularized logistic regression, Random Forest, and eXtreme Gradient Boosting) to detect causal language. A total of 475 sentences, as determined pragmatically based on annotation feasibility and observed stabilization of model performance, were manually labeled as causal or non-causal following established guidelines for systematic evaluation of causal language in observational health research. Of these, 75% of sentences were used for training and 25% for testing. L1-regularized logistic regression achieved the highest performance (accuracy 76%, F1 72%, prevalence detection accuracy 95%, sensitivity 72%, and specificity 80%) and was applied to 20,022 human gut microbiome abstracts published between 2015 and 2025 grouped into 20 thematic topics using structural topic modeling. Predicted causal language prevalence declined from 52% to 44% between 2015 and 2018, then rose to 51% by 2025, with notable variation across topics (range: 43.1-53.3%). Temporal trends differed across subfields, with increases in Metabolic disorders, Fecal microbiota transplantation, and decreases in Biomarkers and prediction, Antibiotic resistance, and In vitro fermentation. Analysis of influential words confirmed that causal meaning is primarily driven by verbs and modifiers lexically signaling change or intervention. The proposed approach for identifying causal claims in scientific abstracts enables systematic and automated, scalable assessment of how evidence is framed. Its application to the microbiome field highlighted heterogeneity in the reporting of causal relationships and informing the interpretation of microbiome findings for clinical and public health decision-making.

The ReproSci project retrospectively analyzed the reproducibility of 1006 claims from 400 papers published between 1959 and 2011 in the field of Drosophila immunity. This project attempts to provide a comprehensive assessment, 14 years later, of the replicability of nearly all publications across an entire scientific community in experimental life sciences. We found that 61% of claims were verified, while only 7% were directly challenged (not reproducible), a replicability rate higher than previous assessments. Notably, 24% of claims had never been independently tested and remain unchallenged. We performed experimental validations of a selection of 45 unchallenged claim, that revealed that a significant fraction (38/45) of them is in fact non-reproducible. We also found that high-impact journals and top-ranked institutions are more likely to publish challenged claims. In line with the reproducibility crisis narrative, the rates of both challenged and unchallenged claims increased over time, especially as the field gained popularity. We characterized the uneven distribution of irreproducibility among first and last authors. Surprisingly, irreproducibility rates were similar between PhD students and postdocs, and did not decrease with experience or publication count. However, group leaders, who had prior experience as first authors in another Drosophila immunity team, had lower irreproducibility rates, underscoring the importance of early-career training. Finally, authors with a more exploratory, short-term engagement with the field exhibited slightly higher rates of challenged claims and a markedly higher proportion of unchallenged ones. This systematic, field-wide retrospective study offers meaningful insights into the ongoing discussion on reproducibility in experimental life sciences.

IntroductionReliable, high-quality research is essential to the field of anaesthesiology. Reproducibility and transparency has been investigated in the biomedical domain and in the social sciences, with both lacking to provide necessary information to reproduce the study findings. In this study, we investigated 14 indicators of reproducibility in anaesthesiology research.\n\nMethodsWe used the National Library of Medicine (NLM) catalogue to search for all anaesthesiology journals that are MEDLINE indexed and provided English texts. PubMed was searched with the list of journals to identify all publications from January 1, 2014 to December 31, 2018. We randomly sampled 300 publications that fit the inclusion criteria for our analysis. Data extraction was then conducted in a blinded, duplicate fashion using a pilot-tested Google form.\n\nResultsThe PubMed search of these journals identified 171,441 publications, with 28,310 being within the time frame. From the 300 publications sampled, 296 full-text publications were accessible. Most of the studies did not include materials or protocol availability statements. The majority of publications did not provide a data analysis script statement (121/122, 99% [98% to 100%]) or a preregistration statement (94/122, 77% [72% to 81%]).\n\nConclusionAnaesthesiology research needs to drastically improve indicators of reproducibility and transparency. By making research publically available and improving accessibility to detailed study components, primary research can be reproduced in subsequent studies and help contribute to the development of new practice guidelines.

The ability to predict scientific breakthroughs at scale would accelerate the pace of discovery and improve the efficiency of research investments. Recent advances in artificial intelligence, graph theory, and computing power have provided new ways to pursue this elusive goal. We have identified a common signature within co-citation networks that accurately predicts the occurrence of breakthroughs in medical research, on average more than 5 years in advance of the subsequent publication(s) that announced the discovery. A combination of features produces these diagnostic signals: a burst of papers exploring a novel scientific concept, an unusually high number of very influential papers in specialty journals, and low topical cohesion of the associated content. We analyzed two different periods separated by 20 years to show that the kinetics of breakthrough formation are conserved, suggesting that our approach can be used to predict which topics will produce future transformative discoveries. Significance statementScientific breakthroughs are rare, as is contemporaneous recognition of their initial expression. Faster, more efficient identification of topics likely to produce future breakthroughs would speed scientific and technological progress. We introduce an AI/ML-detected signature in co-citation networks that recognizes such topics up to twelve years before the breakthrough itself occurs. Our findings illustrate how a better understanding of the scientific process may lead to greater scientific returns.

The (semi-)automated screening of publications for diverse quality and transparency criteria is at the core of systematic literature assessment. Typically, the assessment process involves two initial reviewers and one additional reviewer for cases that require reconciliation. Here, we explore to what extent this process can be assisted by Large Language Models (LLMs). Specifically, whether LLMs are capable of assessing responsible research practices (RRPs) in scientific papers in a robust way. We employed proprietary LLMs to assess an initial set of 37 papers across ten RRPs. The same papers were also reviewed by three human reviewers. We iteratively redesigned prompts to increase model accuracy compared to human ratings which we treated as the gold standard. The resulting pipeline was validated on an additional set of 15 papers. We show that LLM accuracy is comparable to single human reviewer performance (90% for LLM vs 86% for a single human reviewer). However, performance strongly depended on the specific RRPs with accuracy ranging from 40% to 100%. LLMs exhibited an affirmative bias, making more errors when practices were not reported in the papers. Overall, we show how such an approach potentially replaces one human reviewer, enabling AI-assisted assessment of research papers. We discuss how dataset imbalances, validation procedures, and implementation time limit the broad applicability of such approaches. Through this, we develop initial guidance on the utility of proprietary LLMs in evidence synthesis.

Disentangling the evolution of the molecular processes and genetic networks that facilitate the emergence of morphological novelties is one of the main objectives in evolutionary developmental biology. Here, we investigated the evolutionary history of a gene regulatory network controlling the development of novel tooth-like feeding-structures in diplogastrid nematodes. Focusing on NHR-1 and NHR-40, the two transcription factors that regulate the morphogenesis of these feeding structures in Pristionchus pacificus, we sought to determine whether they have a similar function in out-group nematode Caenorhabditis elegans, which has typical "rhabditid" flaps instead of teeth. Contrary to our initial expectations, we found that they do not have a similar function. While both receptors are co-expressed in the tissues that produce the feeding structures in the two nematodes, genetic inactivation of either receptor had no impact on feeding-structure morphogenesis in C. elegans. Transcriptomic experiments revealed that NHR-1 and NHR-40 have highly species-specific regulatory targets. These results suggest two possible evolutionary scenarios: either the genetic module responsible for feeding-structure morphogenesis in Diplogastridae already existed in the last common ancestor of C. elegans and P. pacificus, and subsequently disintegrated in the former as NHR-1 and NHR-40 acquired new targets, or it evolved in conjunction with teeth in Diplogastridae. These findings indicate that feeding-structure morphogenesis is regulated by different genetic programs in P. pacificus and C. elegans, hinting at developmental systems drift during the flap-to-tooth transformation. Further research in other "rhabditid" species is needed to fully reconstruct the developmental genetic changes which facilitated the evolution of novel feeding structures in Diplogastridae. Research HighlightsCombining CRISPR-based mutagenesis, geometric morphometrics, and transcriptomics, we found that the genetic module governing the morphogenesis of novel feeding structures in diplogastrid nematodes is not conserved in the "rhabditid" C. elegans.

Predation can induce behavioral changes in prey, yet the molecular and neuronal mechanisms underlying prey responses remain poorly understood. Here, we investigated how the nematode Caenorhabditis elegans responds to predation by the nematode-trapping fungus, Arthrobotrys oligospora. We found that A. oligospora predation induced quiescence in C. elegans showing rapid cessation of pharyngeal pumping and movement. Calcium imaging revealed that this quiescence was regulated by the activation of sleep-promoting neurons, ALA and RIS. Genetic analyses demonstrated that ALA were essential for pharyngeal pumping inhibition, whereas both ALA and RIS contributed to movement cessation. Transcriptomic analysis in C. elegans showed the upregulation of immune defense genes in response to A. oligospora predation. We demonstrated that mechanosensation was required for pumping inhibition and transcriptomic regulation upon A. oligospora trapping. These findings suggest that physical constraints imposed by fungal traps trigger a stress-induced quiescence and the upregulation of defense genes in C. elegans. We suggest that trapping-induced quiescence might be a predation strategy used by sessile predators to prevail in the evolutionary arms race.

Extracellular vesicle (EV) research has experienced rapid growth in the past decade. While community-driven efforts, such as the MISEV guidelines (2014, 2018, 2023), have aimed to standardize reporting and enhance experimental rigor, these have not yet been universally adopted in daily EV research practices. The complexity and time commitment required by existing reporting tools can deter researchers from fully embracing them, leading to incomplete or inconsistent documentation in EV studies. Therefore, we created EV-checklist, a complementary digital tool to streamline documentation during manuscript submission and increase transparency for the reviewers, editors and readers. Our tool guides researchers through a straightforward checklist covering nomenclature, source, isolation, characterization, and functional studies of EVs. It generates a concise two-page PDF formatted table for easy integration into manuscripts and provides editors, reviewers, and readers with a clear overview of the conducted studies. EV-Checklist intends to complement existing comprehensive registries as a fast-and-easy tool to enhance clarity and accessibility of methodological and functional insights. It may also promote higher adherence to documentation standards and in EV studies. Our tool is available on https://ev-zone.org.