PLOS Biology

Accurate localization of anatomical landmarks is a foundational skill in anatomy and imaging that is often taught informally through expert mentorship, requiring access to data and desktop software. There is no openly accessible, interactive resource that teaches neuroanatomy with quantitative feedback. We present the AFIDs-Validator (validator.afids.io), an open-access, browser-based platform that pairs guided instruction with quantitative assessment. The platform combines (1) a learning mode in which a language-model neuroanatomy tutor operates inside an MRI viewer, giving anatomy-first instruction that responds to the learner's current image slice, orientation, and cursor position; and (2) a validation engine that accepts a learner's landmark file and returns per-landmark Euclidean error against expert-annotated references spanning 21 brain templates. To make the feedback interpretable, we analyzed 15,000 landmark annotations across 132 human subjects and found that landmark difficulty varies fourfold (median error ranged from 0.37 mm at the anterior commissure to 1.50 mm at the temporal horns) with heavy-tailed distributions at every landmark. These distributions are compiled into per-landmark reliability priors, so learners are scored against the empirical spread of trained raters rather than an arbitrary threshold, and difficult landmarks are not mistaken for poor performance. The AFIDs-Validator requires no installation, licensed software, or local data, and all code, reference data, and tutor design are openly released.

Research on butterflies has uncovered a conserved "toolkit" of genes for color pattern development and evolution. One of these genes is optix, a homeobox transcription factor that regulates ommochrome and melanin pigmentation, as well as structural coloration, in nymphalid butterflies. It remains unclear, however, whether optix plays any roles in color patterning outside of the Nymphalidae. We used CRISPR-Cas9 to disrupt optix in the tobacco hornwormmoth Manduca sexta and observed a dramatic abdominal pigmentation phenotype, where orange pigmentation was replaced by black eumelanin. Chemical assays suggest that the orange pigment is not an ommochrome, indicating that optix modulates an alternative, uncharacterized pigment pathway in M. sexta. RNA-seq and chemical analyses of orange and black abdominal scales lead us to speculate that the orange pigment may be a type of melanin, perhaps N-{beta}-alanyldopamine (NBAD) sclerotin. Our results suggest that optix plays a deeply ancestral role in pigment regulation in Lepidoptera, and demonstrates evolutionary flexibility in how it interfaces with pigment chemistry across moths and butterflies.

Western blotting has been widely used for decades, yet figure preparation remains unstandardized, error-prone and difficult to audit. Gelato is a free Fiji/ImageJ plugin that streamlines intuitive figure preparation, logs processing coordinates, provides a straightforward audit platform andenables reproduction from raw images. By making these capabilities simple and accessible, Gelato facilitates robust Western blot reporting, reducing a substantial burden on authors and reviewers while strengthening scientific rigor.

Collateral sensitivity (CS) - resistance to one antibiotic inducing hypersensitivity to another - offers an evolutionary trap for multidrug-resistant pathogens. A strongly connected component (SCC) in the directed CS graph is a closed cycle in which every drug is reachable from every other. Prior evidence for CS SCCs is exclusively in vitro. We mined 104,337 susceptibility records from BV-BRC spanning four WHO critical-priority pathogens (Klebsiella pneumoniae, Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa; 18,821 isolates), using Fisher's exact tests with Benjamini-Hochberg FDR correction, Tarjan's algorithm, and permutation testing (n = 1,000). Two species yielded qualifying SCCs. In K. pneumoniae (4,286 isolates), a 3-node SCC - imipenem, meropenem, tetracycline - was detected (empirical p = 0.001); both carbapenem-tetracycline edges are bidirectional (OR = 1.81-1.82, q < 0.002, n > 850 per edge). In E. coli (6,720 isolates), a bidirectional 2-node SCC links colistin and cefotaxime (OR = 10.13, 95% CI 2.82-46.12, q = 0.042, n = 87; permutation p = 0.008); with a fragility index of 1, we report it as a hypothesis, not an established effect size. The carbapenem signal is tetracycline-specific: tigecycline shows co-resistance (OR < 0.35), as its distinct RamA/AcrAB-TolC mechanism predicts. ORs of 2.2-2.7 persisted across independent year bands (2009-2014). S. aureus returned no qualifying SCC, but that null is power-limited: only 8% of testable pairs could detect the K. pneumoniae effect size. Prior clinical analyses characterised pairwise and three-way collateral effects; to our knowledge these are the first closed CS cycles identified in clinical surveillance data, motivating experimental follow-up.

The archaeon Haloferax volcanii is a unique model organism to investigate replication dynamics and dissect the interplay between multiple origins, alternative replication pathways, and polyploidy. The combination of variable chromosome copy number and asynchronous replication across the cell population gives rise to pronounced phenotypic heterogeneity, further increasing system complexity. To investigate these mechanisms, we implemented (spt-)PALM and STORM microscopy to probe replication dynamics in different cell lines and growth conditions. We established the first STORM-based super-resolution imaging approach of a protein in Haloferax volcanii. Altogether, our results support a model in which each H. volcanii cell replicates a constant fraction of its chromosome copies, irrespective of total ploidy. Consequently, cells with a higher number of chromosomes exhibit proportionally more replication forks, resulting in pronounced ploidy heterogeneity across the population. Moreover, we show that replication foci are not randomly distributed but cluster into discrete Replication Domains containing multiple active sites. This organization persists across different growth conditions and is independent of canonical replication origins, as demonstrated in a strain lacking all four origins. Our findings suggest that replication coordination in H. volcanii is governed by spatial architecture rather than origin-dependent initiation, providing new insights into the evolutionary diversity of replication control mechanisms.

Following an ancestral nocturnal phase, diurnality evolved independently in multiple mammalian lineages after the Cretaceous-Palaeogene boundary. Whether these recurrent temporal-niche transitions drove convergent molecular evolution in the circadian oscillator remains largely untested. We analyzed 18 core clock genes and direct regulators across 60 diurnal and nocturnal mammals, identifying 10 independent gains of diurnality and 5 reversals to nocturnality on a fixed supertree topology. Using six complementary evolutionary frameworks (spanning substitution counts, profile shifts, selective regimes, and evolutionary rates) we tested for shared molecular adaptations. To ensure that our negative findings reflected biological reality rather than methodological insensitivity, we planted 90 synthetic convergent residues into the empirical alignments. While this benchmark demonstrated high power to detect broadly shared convergence (e.g., PCOC recovered 35 of 36 planted sites across ten lineages without false positives), empirical alignments showed no credible signal. Same-residue convergence occurred at expected background levels (0.310 observed vs. 0.337 null), and no gene displayed activity-dependent shifts in selective pressure or evolutionary rate. These results indicate that repeated transitions in mammalian activity patterns were not driven by a common set of detectable amino-acid changes in the core circadian machinery.

Complex syntax is the widely accepted demarcation between animal and human communication. Although animal vocal systems occasionally contain pockets of syntax-like structures, their grammars are thought to be at best regular, whereas human language is supra-regular, i.e., including at least context-free computational complexity. Leveraging a new likelihood method for probabilistic formal grammars, we show that the call sequences of three phylogenetically diverse nonhuman primates (bonobos, olive colobus monkeys, common marmosets) include parts that are most likely generated by supra-regular recursive grammars, similar to human language. These findings challenge syntax as the major transition in hominid evolution and opens up prospects for alternative evolutionary scenarios.

Compositional syntax - the ability to combine meaningful units into structured sequences whose overall meaning depends on both the meanings of the parts and their combination - is a hallmark of human language. Recent work suggests that chimpanzees may exhibit similar capacities, raising the possibility that compositionality predates the emergence of language in humans. However, a key challenge has been to exclude simpler explanations in which call combinations are processed as independent utterances rather than integrated structures. Here, we address this issue using a source-manipulation paradigm in playback experiments with captive chimpanzees (n = 6) at the Sigean African Reserve, France. We presented sequences combining alarm and recruitment calls either from a single source (one loudspeaker) or from two spatially distinct sources (two loudspeakers). If sequences are interpreted as independent calls, responses should not differ between conditions. Instead, chimpanzees responded significantly faster when sequences originated from a single source. Bayesian survival analyses confirmed this effect. These results support a non-trivial form of compositionality, whereby the meaning of call combinations depends on their integration into a unified sequence produced by a single individual. Our findings provide experimental evidence that key features of linguistic compositionality may have evolutionary roots in the primate lineage.

Gene loss is a widespread phenomenon that shapes genome evolution, yet the factors determining why certain genes are repeatedly lost while other functionally related genes are retained remain poorly understood. We addressed this question using the peptide-processing metallocarboxypeptidases carboxypeptidase E (CPE) and carboxypeptidase D (CPD), conserved paralogues that are essential for neuropeptide maturation but strikingly differ in their evolutionary fate: the cpe gene has been independently lost in two major insect lineages, whereas cpd/svr has been universally retained. Combining gene phylogenetic analyses and functional genetics in the beetle Tribolium castaneum, and cross-species rescue experiments in the fly Drosophila melanogaster, we show that CPE and CPD retained partially interchangeable enzymatic functions despite considerable differences in structure, organismal importance and expression. Contrary to expectations, cpe proved more critical than cpd/svr for survival and developmental robustness in Tribolium, while simultaneous RNAi-mediated downregulation of both genes caused complete larval lethality, demonstrating only partial functional redundancy. Moreover, beetle CPE partially rescued the lethal loss of Drosophila CPD, establishing conserved molecular interchangeability across [~]300 million years of insect evolution. Gene phylogenetic analyses further indicate that bilaterian CPE originated through duplication of the second catalytic domain of an ancestral CPD. Together, our results demonstrate that repeated loss of insect cpe cannot be explained by reduced functional importance. Instead, we propose that the structural versatility, broader tissue distribution and multifunctionality of CPD, including its multidomain architecture and splice isoforms, enabled compensation for CPE after gene loss, thereby shaping long-term patterns of gene retention and loss during insect evolution.

Major transitions in evolution are associated with the loss of independent reproduction by formerly autonomous units. Termites provide a powerful system for studying this process because they exhibit diverse social systems in which worker developmental and reproductive potential declines with increasing colony-level organismality. However, the evolutionary sequence and developmental genetic basis of these transitions remain unresolved. Here, using comparative developmental transcriptomics across seven termite species that differ in workers' reproductive potential, we reconstructed the evolutionary history of termite social systems. We found that linear caste development, in which workers retain full reproductive potential, represents the ancestral state of termites. Bifurcated caste development, in which workers partially lose reproductive potential early in development, evolved independently multiple times, with two origins subsequently giving rise to superorganisms with unipotent, sterile workers. Ancestral gene regulatory network (GRN) reconstruction revealed that linear caste development evolved through retention of a juvenile-like worker state and co-option of a conserved developmental GRN characterizing hemimetabolous insect nymphal development, in which juvenile hormone, ecdysone and TGF-{beta} signaling pathways play central roles. The convergent evolution of bifurcated caste development repeatedly co-opted the GRN underlying linear caste development, heterochronically shifting its activity to earlier developmental stages. Finally, we found that the evolution of termite superorganisms involved somatization of the worker caste and co-option of a conserved endocrine GRN for terminal differentiation. Together, these findings uncovered repeated routes to reduced workers' reproductive potential through GRN co-option and highlight striking parallels between superorganism evolution in social insects and organismal evolution in metazoans.

In the wild, a mouse must flexibly perform multiple computations at once, rapidly navigating, sampling its environment, forming and using memories, and balancing internal needs. Capturing the self-paced, low-repetition, complex nature of decision making in the wild while preserving the experimental control needed to interpret its process remains a challenge. Here we show that lab mice, without food restriction, learn to forage for hidden resources cued by ambiguous sounds in a dynamic environment by spatiotemporally reorganizing their existing behavioral repertoire. Using a novel closed-loop assay, we required mice to sit still to receive informative but hard-to-localize sound cues emanating from one correct location they must find, out of 157 possible locations in a large arena. With experience, mice become efficient hunters. Efficient hunting cannot be explained by an increase in a particular behavioral module, but instead, mice reorganize existing behaviors into clustered bouts of high-quality sampling and site-checking. Looking forward, this work establishes an approach for studying the neural, molecular, and evolutionary basis of naturalistic decision-making in mice.

Longitudinal animal experiments generate behavior that is individual, history-dependent, and sometimes affected by ordinary procedural irregularities, yet analyses commonly reduce such records to pooled averages or synchronous trial-level explanations. We introduce history-structured forecasting as an auditable framework for determining whether an animals own preceding behavior carries predictive information beyond current-trial context. We applied the framework to 213,990 events from rats performing an auditory duration-discrimination task, using leakage-safe chronological forward-chaining, explicit trivial baselines, and controls that reset, exchange, or disrupt behavioral history. A transparent gradient-boosted model achieved 51.7% four-class accuracy, exceeding last-action persistence (40.9%) and prefix-derived subject-modal prediction (37.2%); decline-class AUPRC was 0.525 against a prevalence baseline of 0.303. Validation showed that the predictive advantage depended predominantly on each animals short-range sequential action history rather than group-level history, subject identity alone, or reward/correctness features, and strengthened on genuine choice trials. Forecasting remained informative across all 23 labeled animals, including six with recoverable records affected by incorrect training programming. These results revise the interpretation of rewarded give-up behavior while demonstrating how recoverable irregular records can be retained in transparent robustness analyses. History-structured forecasting offers a reusable open-science strategy for extracting reproducible evidence from imperfect longitudinal animal records without creating an artificially clean cohort.

How the brain produces different behaviors from a fixed anatomical substrate is a foundational question in neuroscience. The dominant account holds that behavioral flexibility requires distinct neural circuits: dedicated systems selectively recruited for different processing modes. An alternative proposes that flexibility arises not from switching between circuits, but from dynamically controlling the speed at which a shared circuit operates. Here, we test this account directly using intracranial EEG recordings from nineteen human patients performing two visual search tasks of markedly different difficulty: an easy color-singleton search and a difficult orientation search, a fast vs. slow dissociation attributed for four decades to distinct anatomical and oscillatory neural systems. We show that both search tasks recruit the same cortical regions, oscillatory frequencies, and processing hierarchy. Dynamic time warping reveals that the two neural trajectories are time-stretched versions of one another; the same temporal scaling appears within each condition between faster and slower trials, indexing response speed rather than the distinction between tasks. Alpha-band oscillations (~8-10 Hz) support this temporal scaling through two coordinated mechanisms: synchronizing activity across a posterior cortical network and phase-gating local high-frequency activity (70-150 Hz) within those regions at the same preferred phase in both conditions. These findings establish that, for the two tasks studied here, fast and slow visual search are not implemented by selecting between distinct dedicated circuits, but by adjusting the speed of a single alpha-coordinated architecture. The same principle may apply to other search tasks, and to other domains in which dual-process architectures have been invoked, a testable prediction for future studies.

Alarm pheromone is a high-priority social signal in honey bees, yet direct evidence for how its major component, isopentyl acetate (IPA), is encoded in antennal lobe remains limited. Here, we combine intracellular recording, neuronal staining, and three-dimensional reconstruction to examine neural responses to IPA in the honey bee brain. Integrated analysis of antennal lobe neurons revealed clear but heterogeneous time-locked responses to IPA, which could be grouped into four temporal response motifs: fast transient, monophasic, biphasic excitation-inhibition, and delayed excitation-inhibition. A morphologically identified antennal lobe neuron exhibited a stable excitatory response characterized by short latency and prolonged elevated firing after stimulus onset. In a representative delayed-type antennal lobe neuron, response magnitude showed strong concentration dependence: peak amplitude and post-peak inhibition increased significantly with increasing IPA concentration, whereas peak latency remained largely unchanged. Repeated stimulation at an intermediate concentration produced comparatively modest effects, expressed mainly as attenuation of peak amplitude and a gradual delay in response timing. In addition to antennal lobe neurons, we identified two IPA-responsive protocerebral neurons. Together, these results provide direct single-neuron evidence that IPA is heterogeneously encoded in the honey bee antennal lobe.

The timing of the origins of the crown-group placental mammals has provided one of the classic battlefields in the long-running debate over when clades arise. Undoubted fossil crown-group placentals appear only in the Paleogene, but even so most molecular analyses, and many palaeontologists, have suggested their true origin is somewhere between 70-100 Ma. However, apart from the fact of the fossil record itself, there are several reasons to believe that the true origin is indeed post-Cretaceous, including consideration of the dynamics of stem and crown groups, which strongly favour crown-group origins to lie just after, and not just before, mass extinctions. Here we consider this "hard explosive" model in the light of the newly-developed "Covariant Evolutionary Tempo (CET)" model which allows diversification and molecular evolution rates to covary. It predicts "early bursts" in both lineage creation and molecular evolution at the base of major radiations which lead to highly unequally-sized clades; and an inheritance of rapid rates from this initial event by extant rapidly-evolving clades. We show that when the placentals are constrained to emerge after the K-Pg boundary, they indeed show elevated rates of both diversification and molecular evolution, which rapidly decline. Nevertheless, although elevated, these rates are comparable to the fastest rates seen in extant clades such as the rodents. In addition, the contiguous lineages leading from the origin to the rodents and other fast evolving clades also show elevated rates. These patterns suggest that not only is a Paleogene origin for the placental crown-group plausible, as fossil evidence suggests, but they also provide support for the CET model, which should be considered in other cases of pronounced fossil record/molecular clock mismatch.

Infectious disease transmission models are governed by parameters informed by systematic review of epidemiological literature. Large language models (LLMs) could facilitate this, but the reliability of the results to inform models has not been tested. We built an open-source, end-to-end pipeline to simulate LLM performance in a hypothetical scenario where they were available to inform COVID-19 models developed during the first four months of 2020. It screened articles and extracted the reproduction number, serial interval, and incubation period from full-text PDFs. We applied it to 2,067 PubMed/medRxiv records published 31 December 2019-30 April 2020 using four models (GPT-5-mini, GPT-5.4, Claude Opus 4.8 and Gemini 2.5 Pro) and evaluated it against full-corpus human screening and 50-article extraction gold standards. We combined models post hoc, pooled the extracted values into an infectious disease (SEIR) model, and ran sensitivity analyses to test approaches to improve extraction accuracy. Screening sensitivity was 0.72-0.95 and specificity 0.92-0.99. For articles reporting few values, extraction F1 was 0.90-0.96 with precision 0.91-1.00; across all articles, including those with dozens of stratified estimates, recall fell to 0.38-0.78. No fabricated values observed; errors were misassignments of values filed under the wrong parameter, or borrowed values treated as the study's own. Incomplete extraction from dense articles was mainly due to prompting and output format, not model capability. Ensembling allowed recall-precision trade-offs, and correctness increased with model agreement, from about 30% at one vote to 92-95% at four. The pipeline processed the corpus in hours versus an estimated 130-265 person-hours of manual effort. Our results show that current models can extract transmission parameters from unstructured literature accurately enough to inform outbreak modelling. The bottleneck lies in task specification, and careful prompt and output schema design are key to reducing misassignment errors. Human effort is best directed at workflow development and provenance validation.

Venom systems are key evolutionary innovations that have contributed to the ecological success of numerous animal lineages. In ants (Formicidae), venom exhibits an exceptional diversity in both chemical composition and delivery mechanisms. Here, we combine a synthesis of current knowledge on the evolution of ant venoms with new phylogenetic comparative analyses, integrating chemical, morphological, and phylogenetic data at the genus level. Comparative analyses across Formicidae support the hypothesis that peptide-rich venoms represent the ancestral condition in ants, as they are widely distributed across basal and derived lineages. Despite this ancestral state, we identify seven independent evolutionary transitions toward non-peptidic venom, occurring in major clades including Dolichoderinae, Formicinae, Stenammini, Solenopsidini, Crematogaster, Pheidole, and fungus-growing ants. Ancestral state reconstruction and tests of correlated evolution across 114 genera show that these seven shifts away from peptide-based venom were almost irreversible and tightly coupled with the loss of stinging capacity and a dietary shift away from predation. These transitions are consistently associated with profound morphological modifications of the venom apparatus, such as reduction or complete loss of the sting, or its transformation into non-injecting structures. Chemically, these systems are dominated by alkaloids, iridoids, formic acid, esters, and terpenoids. These repeated transitions reflect major ecological and evolutionary shifts. Lineages that rely less on individual predation and more on collective foraging or plant-based diets tend to exhibit reduced reliance on injectable peptide toxins. Instead, they favor metabolite-based venoms that can be co-opted from existing biochemical pathways and deployed in diverse ecological contexts. Conversely, peptide-rich venoms are retained in lineages subjected to strong selective pressures for rapid prey immobilization or defense against vertebrate predators. Overall, ants represent a unique model for understanding venom evolution, illustrating how shifts in chemistry, morphology, and ecological strategy can drive repeated and convergent innovations. Future integrative studies combining genomics, metabolomics, and functional assays will be essential to unravel the mechanisms underlying these transitions and their role in the extraordinary diversification of ants.

The evolutionary origins of cognitive monitoring remain contested. Metacognition, the capacity to monitor one's own cognitive states, has long been linked to complex vertebrate brains, yet whether a miniature brain can evaluate the reliability of its internal representations is unknown. Here we demonstrate that bumble bees engage in uncertainty-guided decision-making. Bees dynamically adjusted opt-out choices according to perceptual difficulty, settling for a smaller guaranteed reward to avoid errors, and actively paid a reward cost to seek information under uncertainty. Without any retraining, bees transferred this 'opt-out-under-uncertainty' rule to novel tactile and working-memory tasks under an all-probe design. This strategy was stable across individuals and accurately captured by a confidence-based decision model. Our findings suggest that a brain of approximately one million neurons can support an abstract domain-general uncertainty-monitoring policy, indicating that the neural substrates for uncertainty-guided decision-making may be far more ancient and widely distributed than previously assumed.

A fundamental challenge in drug development is the frequent failure of early laboratory research to translate into clinical benefit. One promising solution is to confirm findings from exploratory single-laboratory studies across multiple laboratories before clinical testing. We investigated this approach following the conduct of preclinical multi-laboratory studies across different fields of medicine. For this, we evaluated effect sizes, experimental rigor, and a set of criteria to identify determinants of confirmation success. When tested under increased rigor, only a fraction of multi-laboratory studies confirmed the initial results. The underlying effect size reduction was associated with outcome-relevant experimental differences between exploratory and confirmatory stages. In summary, multi-laboratory studies proved highly informative and served as an effective filter for promising treatments.

Discrete voluntary movement depends on rapid changes in neural excitability across cortical and corticospinal circuits, yet how intrinsic movement-related neural states shape multi-level excitability remains unclear. Here, we combined individualized, state-targeted transcranial magnetic stimulation (TMS) with electroencephalography and electromyography recordings during visually cued finger movements to probe excitability across movement-related beta-band dynamics during two complementary experiments. Immediate transsynaptic cortical excitability closely tracked intrinsic beta dynamics, with attenuation of the second immediate TMS-evoked potential during beta desynchronization and recovery during the post-movement beta rebound. In contrast, corticospinal excitability showed the opposite pattern, with larger motor-evoked potentials during beta desynchronization and reduced responses during beta rebound. Together, these findings identify endogenous beta-state dynamics as a key regulator of movement-related cortical excitability and reveal a fundamental dissociation between how intrinsic brain activity tunes local cortical excitability and corticospinal output in humans.