App-based self-guided mindfulness training for adults with epilepsy: a six-week single-arm feasibility study.

To investigate longitudinal trends in the American Board of Emergency Medicine (ABEM) Qualifying Examination (QE) performance, identify statistically significant inflection points in pass rates, and determine factors associated with the recent decline observed in 2022. This was a retrospective, observational study conducted in 2 phases. Phase 1 analyzed all QE attempts from 1980 to 2024 (n = 83,254) to identify changepoints using segmented regression and Bayes factor comparisons. Phase 2 focused on QE attempts surrounding the most recent changepoint from 2016 to 2024 (n = 23,784) and used multilevel nonlinear piecewise growth models to examine the association between physician characteristics and QE pass rates before and after the 2022 changepoint. Three statistically significant changepoints were identified: 1988, 2003, and 2022. Although increases in pass rates were noted in 1988 and 2003, 2022 was marked by a decline in pass rates with a steeper drop post-2022 (slope from 2016 to 2022 = -0.17 vs slope 2022 to 2024 = -0.67). From 2016 to 2022, MD graduates had a lesser decline in performance compared with non-MD peers. After 2022, those trained during the COVID pandemic also experienced a lesser decline. ABEM QE pass rates experienced 3 major inflection points since 1980, with the most recent occurring in 2022 and representing a meaningful decline. Physician-level characteristics, particularly medical school degree type and COVID-era training, were significantly associated with this trend.

Effects of low-load blood flow restriction vs. high-load resistance training on upper-body strength in male collegiate gymnasts: A randomized controlled trial.

In major depressive disorder (MDD), it is unclear whether fatigue is a result of poor sleep or depression, and its biological basis is unknown. To explore this, Magnetic Resonance Spectroscopy (1H-MRS) in the anterior cingulate cortex (ACC) was acquired in unmedicated participants with MDD (N=76). Gamma-aminobutyric acid (GABA+) and Glx (combination of glutamate and glutamine) concentrations were quantified from an average co-edited difference spectrum, with preprocessing using Gannet and spectral fitting using TARQUIN. Participants (median age: 23.5±14.2years) completed sleep and depression questionnaires. The Wilcoxon rank sum or Kruskal-Wallis test was used to examine the marginal difference in continuous variables between levels of fatigue. Mediation effects of depression severity, GABA+, and Glx concentrations on the relationship between sleep duration and fatigue were evaluated using multivariable logistic regression models with fatigue as the dependent variable. Bayes Factor hypothesis testing examined the strength of evidence. Separate analyses were repeated with sleep efficiency as the main predictor. No significant relationship was observed between sleep duration/efficiency or GABA+/Glx and fatigue. Depression severity was significantly associated with fatigue and had a significant mediation effect on the relationship between sleep duration and fatigue. Specifically, for every additional hour of sleep, the odds of feeling fatigued nearly every day decreased by 11% via an indirect mediating effect of decreased depression severity. GABA+/Glx did not mediate the effects of sleep on fatigue. The results suggest that, in this relatively young population, interventions to improve depression in MDD may be beneficial in reducing the debilitating effects of fatigue.

Inference of interspecific gene flow using genomic data is important to reliable reconstruction of species phylogenies and to our understanding of the speciation process. Gene flow is harder to detect if it involves sister lineages than nonsisters; for example, most heuristic methods based on data summaries are unable to infer gene flow between sisters. Likelihood-based methods can identify introgression between sisters but the test exhibits several nonstandard features, including boundary problems, indeterminate parameters, and multiple routes from the alternative to the null hypotheses. In the Bayesian test, those irregularities pose challenges to the use of the Savage-Dickey (S-D) density ratio to calculate the Bayes factor. Here we develop a theory for applying the S-D approach under nonstandard conditions. We show that the Bayesian test of introgression between sister lineages has low false-positive rates and high power. We discuss issues surrounding the estimation of the rate of gene flow between sister lineages, especially at very low or very high rates, and suggest that evidence for gene flow between sisters be assessed via a Bayesian test. We find that the species split time has a major impact on the information content in the data, with more information at deeper divergence. We use a genomic dataset from Sceloporus lizards to illustrate the test of gene flow between sister lineages.

Evaluating WAIC and PSIS-LOO for bayesian diagnostic classification model selection

Dynamic causal modelling for functional near-infrared spectroscopy using spatial priors derived from diffuse optical tomography.

Interspecific gene flow is commonly inferred using genomic data under the multispecies coalescent (MSC) model. Incomplete taxon sampling can impact inference of gene flow in multiple ways. First unsampled ghost lineages that are sources of introgression may mislead inference of gene flow in analysis of genomic data from sampled species. Second incomplete taxon sampling causes merges of branches on the species phylogeny and complicates the definition and estimation of the rate or magnitude of gene flow, measured by the expected proportion of immigrants in the recipient population (i.e., the introgression probability). We use mathematical analysis and computer simulation to examine the impact of incomplete taxon sampling on inference of gene flow and estimation of its rate using genomic data. We introduce a Bayesian testing approach to select models of gene flow for a species triplet (such as ghost introgression, inflow, and outflow), using the Savage-Dickey density ratio to calculate Bayes factors. We show that the approach has excellent sensitivity and specificity, whereas heuristic methods based on data summaries typically cannot distinguish among those scenarios. We find that genomic data allow reliable estimation of the proportion of immigrants (rather than the number of immigrants), even when the assumed demographic model is incorrect due to incomplete taxon sampling. When population size differs among species, assuming the same size may lead to seriously biased estimates of the rate of gene flow. The f-branch approach is effective in reducing the number of gene-flow events suggested by triplet analyses but often fails to identify the correct model of gene flow and tends to underestimate the rate of gene flow. Our results highlight the need for improving summary methods to accommodate different population sizes and to infer gene flow between sister lineages.

Constraints on the cosmological parameters of Torsion Condensation (TorC) are investigated using Planck 2018 Cosmic Microwave Background data. TorC is a case of Poincaré gauge theory — a formulation of gravity motivated by the gauge field theories underlying fundamental forces in the standard model of particle physics. Unlike general relativity, TorC incorporates intrinsic torsion degrees of freedom while maintaining second-order field equations. At specific parameter values, it reduces to the ΛCDM model, providing a natural extension to standard cosmology. The base model of TorC introduces two parameters beyond those in ΛCDM: the initial value of the torsion scalar field and its time derivative — one can absorb the latter by allowing the dark energy density to float. To constrain these parameters, the PolyChord nested sampling algorithm is employed, interfaced via Cobaya with a modified version of CAMB. Our results indicate that TorC allows for a larger inferred Hubble constant, offering a potential resolution to the Hubble tension. Tension analysis using the R-statistic shows that TorC alleviates the statistical tension between the Planck 2018 and SH0ES 2020 datasets, though this improvement is not sufficient to decisively favour TorC over ΛCDM in a Bayesian model comparison. This study highlights TorC as a compelling theory of gravity, demonstrating its potential to address cosmological tensions and motivating further investigations of extended theories of gravity within a cosmological context. As current and upcoming surveys — including Euclid, Roman Space Telescope, Vera C. Rubin Observatory, LISA, and Simons Observatory — deliver data on gravity across all scales, they will offer critical tests of gravity models like TorC, making the present a pivotal moment for exploring extended theories of gravity.

DarkSide-50 achieved leading WIMP limits down to 1.2 GeV/c 2 with an ionization-only analysis, despite its small active mass of 46 kg compared to multi-ton noble-liquid detectors. Accurate modelling of the nuclear-recoil ionization yield (Qy ) is central to interpreting such searches. A new global fit combining nuclear-recoil response measurements from DarkSide-50, ARIS, SCENE, and the recent ReD experiment constrains Qy between 0.4 and 150 keV within the Thomas-Imel box framework. The dependence on screening potentials is addressed through a Bayesian model comparison, which rejects the ZBL and Molière functions and strongly favours the Lenz-Jensen one. The resulting model improves DarkSide-50 sensitivity below 3 GeV/c 2 and refines the DarkSide-20k sensitivity accordingly.

Identifying appropriate structures for generative or world models is essential for both biological organisms and machines. This work shows that synaptic pruning facilitates efficient statistical structure learning. We extend previously established canonical neural networks to derive a synaptic pruning scheme that is formally equivalent to an online Bayesian model selection. The proposed scheme, termed Bayesian synaptic model pruning (BSyMP), utilizes connectivity parameters to switch between the presence (ON) and absence (OFF) of synaptic connections. Mathematical analyses reveal that these parameters converge to zero for uninformative connections, thus providing reliable and efficient model reduction. This enables the identification of a plausible structure for the environmental model, particularly when the environment is characterized by sparse likelihood and transition matrices. Through causal inference and rule learning simulations, we demonstrate that BSyMP achieves model reduction more efficiently than the conventional Bayesian model reduction scheme. These findings indicate that synaptic pruning could be a neuronal substrate underlying structure learning and generalizability in the brain.

A particularly compelling aspect of the GW190521 event detected by the LIGO-Virgo-KAGRA (LVK) collaboration is that it has an extremely short duration, and lacks a clearly identifiable inspiral phase usually observed in the binary black holes (BBHs) coalescence. In this work, we hypothesize that GW190521 might represent a single, isolated gravitational wave (GW) echo pulse from the wormhole, which is the postmerger remnant of BBHs in another universe and connected to our universe through a throat. The ringdown signal after BBHs merged in another universe can pass through the throat of wormhole and be detected in our universe as a short-duration echo pulse. Our analysis results indicate that our model yields a network signal-to-noise ratio comparable to that of the standard BBHs merger model reported by the LVK collaboration. For GW190521, Bayesian model selection yields ln ℬEcho BBH ≃ -2.9, indicating that the data favor the BBH hypothesis over our echo-for-wormhole model.

The Bayesian analysis of infectious disease surveillance data from multiple locations typically involves building and fitting a spatio-temporal model of how the disease spreads in the structured population. Here we present new generally applicable methodology to perform this task. We introduce a parsimonious representation of seasonality and a biologically informed specification of the outbreak component to avoid parameter identifiability issues. We develop a computationally efficient Bayesian inference methodology for the proposed models, including techniques to detect outbreaks by computing marginal posterior probabilities at each spatial location and time point. We show that it is possible to efficiently integrate out the discrete parameters associated with outbreak states, enabling the use of dynamic Hamiltonian Monte Carlo (HMC) as a complementary alternative to a hybrid Markov chain Monte Carlo (MCMC) algorithm. Furthermore, we introduce a robust Bayesian model comparison framework based on importance sampling to approximate model evidence in high-dimensional space. The performance of our methodology is validated through systematic simulation studies, where simulated outbreaks were successfully detected, and our model comparison strategy demonstrates strong reliability. We also apply our new methodology to monthly incidence data on invasive meningococcal disease from 28 European countries. The results highlight outbreaks across multiple countries and months, with model comparison analysis showing that the new specification outperforms previous approaches. The accompanying software is freely available as a R package at https://github.com/Matthewadeoye/DetectOutbreaks.

In the past decade, there has been considerable concern about the reproducibility of psychological research. We suggest that in-class replications are a feasible setting for the replication of psychophysiological findings and present our conceptual in-class replication of a finding by Sussman, Ritter, and Vaughan Jr. (1998). In the original study, infrequent pitch deviants (proportion: 20%) were presented either at fixed (i.e., predictable) positions or at random positions (i.e., as classic oddballs) in the auditory sequence. The authors found that deviants presented in the predictable condition did not elicit the mismatch negativity (MMN) in the event-related potential (ERP) when the tone sequence was presented at sufficiently fast stimulus onset asynchronies (SOAs) but deviants presented in the random condition elicited an MMN. We replicated a subset of the original conditions in a sample of 25 participants. Although we had changed some aspects of the original design we found frequentist and Bayesian statistical evidence in favor of ERP-differences in the predictable vs. the random condition in the MMN time window. In line with the original results deviants presented in the predictable condition did not elicit an MMN suggesting that the auditory system extracts regularly occurring patterns within fast-paced task-irrelevant acoustic input. However, sequential Bayes factor analysis showed a substantial between-participant variability in the random condition obscuring the presence of a reliable MMN. We discuss potential inter-individual differences in segregating fast-paced sound sequences into separate streams as an explanation for this phenomenon. We further discuss advantages and disadvantages of in-class replications in psychophysiological research.

Abstract Our identities are composed of multiple categories—such as ethnicity, gender, and socioeconomic status—that shape our social networks through group-based connection preferences. As most research focuses on one-dimensional interactions, a fundamental question remains: How do individuals integrate multidimensional identity information when forming social ties? Addressing this question is crucial for understanding social dynamics, mitigating segregation, and designing integration-promoting interventions. Here, we develop a principled theoretical framework to model multidimensional social interactions. Using Bayesian model selection, we compare competing preference aggregation mechanisms in two empirical systems (high-school friendship networks and marriages) and find that a simple latent preference model outperforms more complex alternatives: people evaluate each identity dimension independently and form ties mainly when all evaluations are favorable. The model yields interpretable preference estimates and principled measures of dimension salience, providing a practical tool for analyzing social choices and identifying which aspects of identity matter most for tie formation. Our work reveals how social structures emerge from intersecting identities, with broad implications for understanding social cohesion and addressing intersectional inequalities in networks.

Abstract Yukawa gravity provides a generalized framework for modeling gravity modification. We investigate the rotation curve profiles of spiral galaxies under Yukawa-like theories governed by the coupling strength $$\beta $$ β and the interaction range $$\lambda $$ λ . We develop a unified analytical and numerical framework to calculate rotational velocities under Yukawa gravity, which includes contributions from all major galactic components: stellar bulge, disk, dark matter (DM) halo, and central supermassive black hole. The calculations show that $$\beta $$ β and $$\lambda $$ λ strongly influence velocity distributions by shifting peaks, creating double-peaked structures, or enhancing dark matter dominance in the bulge or disk. To assess observational implications, we perform Bayesian analyses using data from the Milky Way (MW) and Andromeda (M31), which offer complementary characteristics: MW provides precise velocity profiles across multiple scales, while M31 includes broader morphological constraints. We examine four scenarios: Yukawa gravity without dark matter, dark matter with non-trivial coupling, fully modified gravity, and standard Newtonian gravity. Results show that MW models with $$\lambda < 1$$ λ < 1 kpc yield high Bayes factors but risk overfitting, as dark matter mimics baryonic kinematics, while M31’s photometric priors from conjugate observations mitigate this, yielding robust parameter estimates. However, in M31, Bayes factors favor Newtonian gravity, suggesting that current data lack the precision to resolve more complex models. This finding highlights two key needs: (i) realistic, physically or empirically informed priors to avoid biased constraints, and (ii) high-precision data with independent photometry to guard against overfitting. Our framework offers a scalable approach for testing gravity with large galactic rotation curve datasets.

Flexible Bayesian Inference for Identifying Significantly Correlated Multiple Pathway Sets.

Abstract This work reexamines cosmological parameter constraints from the DESI Data Release 2 baryon acoustic oscillation (BAO) measurements using the distance-basis representation $$(D_V/r_d,\,D_M/D_H)$$ ( D V / r d , D M / D H ) , which separates the isotropic BAO scale from the scale-free Alcock–Paczynski ratio. This work compares $$\Lambda $$ Λ CDM, $$\omega \textrm{CDM}$$ ω CDM , and $$\omega _{0}\omega _{a}\textrm{CDM}$$ ω 0 ω a CDM models to evaluate how the choice of data basis and the width of the prior on $$\omega _a$$ ω a affect dark-energy inference. Ratio-only fits ( $$D_M/D_H$$ D M / D H ) amplify the $$(\omega _0,\,\omega _a)$$ ( ω 0 , ω a ) degeneracy and can produce large apparent shifts in point estimates without genuine evidence for dynamical dark energy. Joint fits using $$(D_V/r_d, D_M/D_H)$$ ( D V / r d , D M / D H ) restore parameter consistency and show that these shifts mainly trace the degeneracy ridge. The pivoted equation of state, $$\omega _p=\omega (a_p)\simeq -0.9\pm 0.1$$ ω p = ω ( a p ) ≃ - 0.9 ± 0.1 at $$z_p\simeq 0.34$$ z p ≃ 0.34 , remains stable and consistent with a cosmological constant within $$1\sigma $$ 1 σ . Model-selection diagnostics (AIC, BIC, and Bayes factors) provide only moderate support for $$\Lambda $$ Λ CDM, indicating no significant evidence for an evolving $$\omega (a)$$ ω ( a ) . These findings clarify the interplay among basis choice, absolute-scale anchoring, and degeneracy geometry in BAO-only dark-energy analyses, providing a benchmark for future DESI and next-generation surveys.

Generalized Bayesian multidimensional scaling and model comparison

In pulsar timing array (PTA) data analysis, noise is typically assumed to be Gaussian, and the marginalized likelihood has a well-established analytical form derived within the framework of Gaussian processes. However, this Gaussianity assumption may break down for certain classes of astrophysical and cosmological signals, particularly for a gravitational wave background (GWB) generated by a population of supermassive black hole binaries (SMBHBs). In this work, we present a new method for testing the presence of non-Gaussian features in PTA data. We go beyond the Gaussian assumption by modeling the noise or signal statistics using a Gaussian mixture model (GMM). An advantage of this approach is that the marginalization of the likelihood remains fully analytical, expressed as a linear combination of Gaussian PTA likelihoods. This makes the method straightforward to implement within existing data analysis tools. Moreover, this method extends beyond the free spectrum analysis by producing posterior probability distributions of higher-order moments inferred from the data, which can be incorporated into spectral refitting techniques. We validate the model using simulations and demonstrate the sensitivity of PTAs to non-Gaussianity by computing the Bayes factor in favor of the GMM as a function of the injected excess moments. We apply the method to a more astrophysically motivated scenario where a single SMBHB is resolved on top of a Gaussian GWB and show that significant non-Gaussianities are introduced by the individual source. Finally, we test our model on a realistic GWB generated from a simulated population of SMBHBs.