Nevin Manimala

Phys Rev Lett. 2025 May 2;134(17):177301. doi: 10.1103/PhysRevLett.134.177301.

ABSTRACT

Transfer learning (TL) is a well-established machine learning technique to boost the generalization performance on a specific (target) task using information gained from a related (source) task, and it crucially depends on the ability of a network to learn useful features. Leveraging recent analytical progress in the proportional regime of deep learning theory (i.e., the limit where the size of the training set P and the size of the hidden layers N are taken to infinity keeping their ratio α=P/N finite), in this Letter we develop a novel single-instance Franz-Parisi formalism that yields an effective theory for TL in fully connected neural networks. Unlike the (lazy-training) infinite-width limit, where TL is ineffective, we demonstrate that in the proportional limit TL occurs due to a renormalized source-target kernel that quantifies their relatedness and determines whether TL is beneficial for generalization.

PMID:40408730 | DOI:10.1103/PhysRevLett.134.177301

Phys Rev Lett. 2025 May 2;134(17):178201. doi: 10.1103/PhysRevLett.134.178201.

ABSTRACT

Understanding how a fluid turns into an amorphous solid is a fundamental challenge in statistical physics, during which no apparent structural ordering appears. In the athermal limit, the two states are connected by a well-defined jamming transition, near which the solid is marginally stable. A recent mechanical response screening theory proposes an additional transition above jamming, called a plastic-to-elastic transition here, separating anomalous and quasielastic mechanical behavior. Through numerical inflation simulations in two dimensions, we show that the onsets of long-range radial and angular correlations of particle displacements decouple, occurring, respectively, at the jamming and plastic-to-elastic transitions. The latter is characterized by a power-law diverging correlation angle and a power-law spectrum of the displacements along a circle. This work establishes two-step transitions on the mechanical properties during “decompression melting” of an athermal overjammed amorphous solid, reminiscent of the two-step structural melting of a crystal in two dimensions. In contradistinction with the latter, the plastic-to-elastic transition exists also in three dimensions.

PMID:40408728 | DOI:10.1103/PhysRevLett.134.178201

Phys Rev Lett. 2025 May 9;134(18):180405. doi: 10.1103/PhysRevLett.134.180405.

ABSTRACT

The eigenstate thermalization hypothesis (ETH) has been highly influential in explaining thermodynamic behavior of closed quantum systems. As of yet, it is unclear whether and how the ETH applies to non-Hermitian systems. Here, we introduce a framework that extends the ETH to non-Hermitian systems, within which expectation values of local operators reproduce statistical and scaling predictions known from Hermitian ETH. We illustrate the validity of the framework on non-Hermitian random-matrix and Sachdev-Ye-Kitaev models. Further, we show numerically how the static ETH predictions become imprinted onto the dynamics of local observables. Finally, we present a prescription for observing both ETH-obeying and ETH-violating regimes in an optical-lattice experiment that implements a disordered interacting Hatano-Nelson model. Our results generalize the celebrated ETH to the non-Hermitian setting, and they show how it affects the system dynamics, and how the salient signatures can be observed in present-day cold-atom experiments.

PMID:40408674 | DOI:10.1103/PhysRevLett.134.180405

PLoS One. 2025 May 23;20(5):e0321008. doi: 10.1371/journal.pone.0321008. eCollection 2025.

ABSTRACT

Accurate monthly runoff forecasting is vital for water management, flood control, hydropower, and irrigation. In glacierized catchments affected by climate change, runoff is influenced by complex hydrological processes, making precise forecasting even more challenging. To address this, the study focuses on the Lotschental catchment in Switzerland, conducting a comprehensive comparison between deep learning and ensemble-based models. Given the significant autocorrelation in runoff time series data, which may hinder the evaluation of prediction models, a novel statistical method is employed to assess the effectiveness of forecasting models in detecting turning points in the runoff data. The performance of Extreme Gradient Boosting (XGBoost) was compared with long short-term memory (LSTM) and random forest (RF) models for one-month-ahead runoff forecasting. The study used 20 years of runoff data (2002-2021), with 70% (2002-2015) dedicated for training and calibration, and the remaining data (2016-2021) for testing. The findings for the testing phase results show that the XGBoost model achieves the best accuracy, with R² of 0.904, RMSE of 1.554 m³/sec, an NSE of 0.797, and Willmott index (d) of 0.972, outperforming both the LSTM and RF models. The study also found that the XGBoost model estimated turning points more accurately, obtaining forecasting improvements of up to 22% to 34% compared to LSTM and RF models. Overall, the study’s findings are essential for global water resource management, providing insights that can inform sustainable practices to support societies impacted by climate change.

PMID:40408639 | DOI:10.1371/journal.pone.0321008

PLoS One. 2025 May 23;20(5):e0320726. doi: 10.1371/journal.pone.0320726. eCollection 2025.

ABSTRACT

Clinical breast cancer screening among Ghanaian women is generally unsatisfactory due to poor knowledge despite its critical role in the early detection of breast cancer. Available studies in Ghana show that Islamic women have poorer screening rates which may be due to sociocultural and religious barriers. Understanding the factors that influence clinical breast cancer screening among Islamic women is a critical step that can help the design of initiatives to increase screening among Muslim populations in Ghana. Therefore, this study aimed to explore the factors influencing clinical breast cancer screening among Islamic women in the Kumasi Metropolis of Ghana. From August 20, 2024, to November 01, 2024, a community-based cross-sectional systematic sampling technique was deployed in the Aboabo and Asawase communities of the Ashanti Region to select 500 Islamic women for the study. Binary logistic regression was employed to determine the relationships between variables. Outcome variables with P-values < 0.05 were considered statistically significant. Most of the respondents were of Ghanaian Northern ancestry, with secondary-level education as the highest educational attainment. Compared to women with low cultural and religious norms, women with stronger personal and religious norms had 0.61 lower odds of screening (aOR=0.61, 95% CI = 0.34-1.08). Participant’s level of religiosity had a significant association with clinical breast cancer screening, with 1.16 times higher odds of screening (aOR= 1.16, 95% CI = 1.02-1.32) after adjusting for the covariates. Islamic women perceived high benefits of clinical breast cancer screening but fear of personal and social norm violations at the screening centers and poor knowledge about breast cancer limited their actual participation in clinical breast cancer screening practices. Implementing a national breast cancer education campaign to emphasize the need for asymptomatic or routine screening and provider training on culturally competent practices is encouraged.

PMID:40408633 | DOI:10.1371/journal.pone.0320726

PLoS One. 2025 May 23;20(5):e0323717. doi: 10.1371/journal.pone.0323717. eCollection 2025.

ABSTRACT

The escalating obesity epidemic in sub-Saharan Africa is a pressing regional concern. Despite this, there is scarce evidence of effective strategies to halt its upward trend in the region. We have, therefore, synthesised evidence on effective interventions to prevent and manage obesity in sub-Saharan Africa. We searched Scopus, PsycINFO, Cochrane Library, Web of Science and Medline for pertinent studies for this review. Studies were eligible if they focused on a sub-Saharan African country and assessed obesity/overweight with objective outcome measures. We examined their methodological quality with the Joanna Briggs Institute and the National Institutes of Health appraisal checklists. Publication bias was assessed with funnel plots. A meta-analysis with a random-effects model was fitted to explore the pooled effect of identified obesity interventions on anthropometric obesity measures. The heterogeneity of the studies was assessed using the I-square statistic. Our search yielded seven eligible studies for this review. Their quality ranged from moderate to high. The interventions identified included aerobic and resistance exercises, micronutrient supplementation and physical education. The meta-analysis revealed that aerobic and resistance training could significantly reduce obesity by approximately 34% (p = 0.04; 95%CI = -0.67 – -0.02). However, they do not significantly reduce waist circumference (Effect size = -1.14; 95%CI = -0.67-0.55; p = 0.19). Aerobic and resistance training exercises could be embedded in physical activity interventions to prevent and manage overweight and obesity in sub-Saharan Africa. PROSPERO registration number: CRD42023430503.

PMID:40408622 | DOI:10.1371/journal.pone.0323717

PLoS One. 2025 May 23;20(5):e0323733. doi: 10.1371/journal.pone.0323733. eCollection 2025.

ABSTRACT

OBJECTIVES: The high burden of preventable disease among men in many countries has highlighted the urgency of promoting stronger engagement by men in health services and programs. In order to inform prevention and early intervention strategies in Australia, this study aimed to examine how age and other socio-demographic factors moderate help-seeking preferences among men in this population, and the major psychosocial and practical barriers to healthcare use for men across the life course.

DESIGN: Cross-sectional survey using a nationally representative sample.

SETTING: Online survey in March 2021.

PARTICIPANTS: English-speaking Australian men aged 18-years and older, recruited using a probabilistic sampling method. Of the 1,409 men invited to participate, 1,282 (91%) completed the survey.

MAIN OUTCOME MEASURES: Preferences for help-seeking related to physical and mental health, and psychosocial and practical barriers to help-seeking.

RESULTS: Compared to men aged 70 years and over, younger men were less likely to choose professional sources of help, with those aged 18-29 years showing the lowest odds when needing assistance for physical (OR = 0.28, 95% CI 0.17-0.49) and mental health (OR = 0.28, 95% CI 0.16-0.46). Men in this youngest age group also reported experiencing more practical barriers to healthcare access with 77 out of 241 (32%) men experiencing three or more barriers to healthcare engagement, compared to 16 out of 172 (9.3%) men over 70 years. Multivariable analysis showed that younger age was associated with higher psychosocial barriers to help-seeking.

CONCLUSIONS: Age is a significant factor in men’s health help-seeking preferences in Australia and these findings highlight the unique help-seeking profile of younger men. Younger Australian men are less likely to seek help proactively, and encounter more practical and psychosocial barriers than older men. The findings underscore the necessity for public health strategies to engage younger men effectively in proactive health management.

PMID:40408621 | DOI:10.1371/journal.pone.0323733

PLoS One. 2025 May 23;20(5):e0323708. doi: 10.1371/journal.pone.0323708. eCollection 2025.

ABSTRACT

INTRODUCTION: The ongoing coronavirus disease 2019 (COVID-19) pandemic, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), presents a significant global public health threat. Concurrently, hepatitis B virus (HBV) remains a significant public health challenge. While previous studies have indicated an association between COVID-19 and chronic hepatitis B, the common underlying pathogenesis of these diseases remains incompletely understood.

METHODS: To investigate the shared molecular mechanisms between chronic HBV infection and COVID-19, a comprehensive investigation was conducted using bioinformatics and systems biology. Specifically, we utilized RNA-seq datasets (GSE196822 and GSE83148) to identify differentially expressed genes (DEGs) associated with both SARS-CoV-2 and HBV infection. Subsequently, these common DEGs were utilized to identify shared pathways, hub genes, transcriptional regulatory networks, and potential drugs. The differential expression of hub genes in both COVID-19 and HBV was verified using the GSE171110 and GSE94660 datasets, respectively.

RESULTS: From the 106 shared DEGs identified, immune-related pathways were found to play a role in the development and progression of chronic hepatitis B and COVID-19. Protein-protein interaction (PPI) network analysis revealed 8 hub genes: CDK1, E2F7, E2F8, TYMS, KIF20A, CENPE, TPX2, HMMR, CD8A, GZMA. In the validation set, the expression of hub genes was statistically significant in both the COVID-19 group and the HBV group compared with the healthy control group. Transcriptional regulatory network analysis identified 155 microRNAs (miRNAs) and 43 transcription factors (TFs) as potential regulatory signals. Notably, we identified potential therapeutic drugs for HBV chronic infection and COVID-19, including progesterone, estradiol, dasatinib, aspirin, etoposide, irinotecan hydrochloride, phorbol 12-myristate 13-acetate, lucanthone, calcitriol.

CONCLUSION: This research elucidates potential molecular targets, signaling pathways, and promising small molecule compounds that could aid in the treatment of chronic HBV infection and COVID-19.

PMID:40408617 | DOI:10.1371/journal.pone.0323708

JCO Clin Cancer Inform. 2025 May;9:e2500046. doi: 10.1200/CCI-25-00046. Epub 2025 May 23.

ABSTRACT

PURPOSE: Many trials ask patients to complete patient-reported outcome measures (PROMs) via the web, excluding patients unable to use/access the Internet. The PRO-TECT trial (AFT-39, ClinicalTrials.gov identifier: NCT03249090) also offered a telephone interface option (interactive voice response [IVR]). We compared patients choosing IVR versus web on alert rates to nurses and clinical outcomes to determine if a telephone option can close disparities in symptom management.

METHODS: PRO-TECT randomized 26 community oncology practices to the PROM intervention arm where concerning symptoms generated automated alerts to nurses. IVR and web patients were compared for social determinants of health (SDOH) using analysis of variance and chi-square tests. After accounting for clustering and confounders, we used generalized estimating equations to compare alert rates, mixed models for quality of life (QOL) at 3 months, and Cox regression for emergency visits and survival at 12 months.

RESULTS: Among 593 patients, 215 (36%) chose IVR and 378 (64%) chose web. IVR patients were older (65.2 v 60.8 years) and were more often rural residents (32% v 23%), Black (27% v 11%), and with less education (54% v 27% ≤high school; all P < .01). Patients choosing IVR had more surveys with concerning symptoms (49% v 37%) and nurses felt clinical attention was warranted more often (4.8 surveys v 3.4 surveys; all P < .001) but ultimately experienced similar benefits as web in QOL, emergency visits, and survival.

CONCLUSION: One third of community patients choose a telephone option over the web for reporting PROMs during cancer care. These patients are disproportionately from SDOH backgrounds at risk of poor clinical outcomes and have higher symptom management needs but ultimately experience similar clinical benefits as patients choosing the web. PROM programs should offer web alternatives to close disparities in symptom management.

PMID:40408610 | DOI:10.1200/CCI-25-00046

PLoS One. 2025 May 23;20(5):e0323736. doi: 10.1371/journal.pone.0323736. eCollection 2025.

ABSTRACT

The COVID-19 pandemic underscored the urgent need for widespread vaccination to achieve herd immunity and mitigate severe outcomes. To address vaccine supply constraints, heterologous prime-boost strategies were adopted in Korea and other countries. Although studies have explored the effectiveness of heterologous prime-boost SARS-CoV-2 vaccination, comprehensive research on its adverse events (AEs), particularly severe serious AEs (SAEs), remains lacking. As an observational study, this study aims to compare severe SAEs across vaccination strategies and examine factors, including heterologous vaccination, associated with 42-day mortality among patients with severe SAEs, without implying causality. Our retrospective cohort study involved 358 cases of severe SAEs following prime-boost SARS-CoV-2 vaccination in Gyeonggi Province, South Korea, from February 26, 2021, to March 15, 2022. In patients with severe SAEs, the heterologous vaccination was associated with a higher risk of mortality than the homologous viral vector vaccination. Vaccinations performed at vaccination centers were associated with a lower risk of mortality. Furthermore, among patients with severe SAEs, the heterologous group exhibited a higher rate of respiratory diseases and genitourinary diseases compared to the homologous viral vector group. Moreover, the rate of deaths from genitourinary diseases among patients with severe SAEs was significantly higher in the heterologous group compared to the homologous viral vector group. We believe that our study, while limited to associations and not establishing causality, provides critical insights that could inform decision-making in scenarios where heterologous vaccination is necessitated by vaccine shortages or other constraints, particularly in managing severe SAEs and improving patient outcomes.

PMID:40408606 | DOI:10.1371/journal.pone.0323736