Category: Statistics

Ann Bot. 2025 Jul 10:mcaf138. doi: 10.1093/aob/mcaf138. Online ahead of print.

ABSTRACT

BACKGROUND AND AIMS: Xylem and vessel structures help to maintain water transport in woody plants in response to environmental changes. Many studies have demonstrated the relationships between vessel structures and, in particular, temperature. However, the effects of environmental factors on the of vessel size distribution and on vessels of different sizes have not been fully assessed statistically. Lianas are characterized by large vessels and vessel dimorphism. Liana abundance decreases with decreasing temperature; this pattern is attributed to the vulnerability of their large vessels to freeze-thaw embolism. However, in temperate zones, the relationships between liana abundance and temperature differ between twining and root climber.

METHODS: We sampled wood discs of eight liana species distributed across Japan and measured size and shape of 130,940 vessels in 836 sections from 219 individuals. We classified vessels of each species into two diameter clusters (large and small) and calculated vessel traits and potential hydraulic conductivity (Kp). Vessel traits were compared among climbing mechanisms and the relationships between vessel traits and temperature were analyzed for species.

KEY RESULTS: Twining climbers had larger vessel diameters than root climbers and had greater Kp. However, the relationships between temperature and vessel traits of species were inconsistent within climbing mechanisms. The decrease in Kp in certain species with decreasing temperature might result from species-specific changes in xylem structure. Vessels of the two clusters related differently to temperature in some species.

CONCLUSIONS: The vessel traits in each climbing mechanism might partly explain the distribution patterns of these lianas in the study region. Furthermore, changes in Kp in some species supported the prediction that liana competitiveness decreases with decreasing temperature. Understanding the mechanisms behind the changes in vessel traits and vessel size distribution along environmental factors will provide fundamental insights into how environmental changes affect forest ecosystems by altering plant hydraulic function.

PMID:40638780 | DOI:10.1093/aob/mcaf138

JMIR Med Inform. 2025 Jul 10;13:e67513. doi: 10.2196/67513.

ABSTRACT

BACKGROUND: Adverse drug reactions (ADRs) pose serious risks to patient health, and effectively predicting and managing them is an important public health challenge. Given the complexity and specificity of biomedical text data, the traditional context-independent word embedding model, Word2Vec, has limitations in fully reflecting the domain specificity of such data. Although Bidirectional Encoder Representations from Transformers (BERT)-based models pretrained on biomedical corpora have demonstrated high performance in ADR-related studies, research using these models to predict previously unknown drug-side effect relationships remains insufficient.

OBJECTIVE: This study proposes a method for predicting drug-side effect relationships by leveraging the parametric knowledge embedded in biomedical BERT models. Through this approach, we predict promising candidates for potential drug-side effect relationships with unknown causal mechanisms by leveraging parametric knowledge from biomedical BERT models and embedding vector similarities of known relationships.

METHODS: We used 158,096 pairs of drug-side effect relationships from the side effect resource (SIDER) database to generate an adjacency matrix and calculate the cosine similarity between word embedding vectors of drugs and side effects. Relation scores were calculated for 8,235,435 drug-side effect pairs using this similarity. To evaluate the prediction accuracy of drug-side effect relationships, the area under the curve (AUC) value was measured using the calculated relation score and 158,096 known drug-side effect relationships from SIDER.

RESULTS: The clagator/biobert_v1.1 model achieved an AUC of 0.915 at an optimal threshold of 0.289, outperforming the existing Word2Vec model with an AUC of 0.848. The BERT-based models pretrained on the biomedical corpus outperformed the vanilla BERT model with an AUC of 0.857. External validation with the FDA (Food and Drug Administration) Adverse Event Reporting System data, using Fisher exact test based on 8,235,435 predicted drug-side effect pairs and 901,361 known relationships, confirmed high statistical significance (P<.001) with an odds ratio of 4.822. In addition, a literature review of predicted drug-side effect relationships not confirmed in the SIDER database revealed that these relationships have been reported in recent studies published after 2016.

CONCLUSIONS: This study introduces a method for extracting drug-side effect relationships embedded in parameters of language models pretrained on biomedical corpora and using this information to predict previously unknown drug-side effect relationships. We found that BERT-based models pretrained with biomedical corpora consider contextual information and achieve better performance in drug-side effect relationship prediction. External validation using the FDA Adverse Event Reporting System dataset and the literature review of certain cases confirmed high statistical significance, demonstrating practical applicability. These results highlight the utility of natural language processing-based approaches for predicting and managing ADR.

PMID:40638775 | DOI:10.2196/67513

PLoS One. 2025 Jul 10;20(7):e0326307. doi: 10.1371/journal.pone.0326307. eCollection 2025.

ABSTRACT

The chronic small-business closure has emerged as a critical economic issue in South Korea, considering more than half of businesses have been closed within three years. While some previous literature has analyzed the causes and their negative impacts on the economy, there is still a lack of studies on understanding the pattern dynamics and predicting future possible closure scenarios due to the lack of appropriate data. Using 3,000,000 individual commercial facility data from 2004 to 2018 in Seoul, Korea, the primary purpose of this research is two-fold: (1) to develop a methodological framework to simulate survival rate pattern change using a deep-learning based model and (2) to reclassify the commercial districts based on the prediction outcomes to 8 survival rate change types. The results indicate that the LSTM model can be useful in predicting and simulating the survival rate of commercial facilities. Moreover, the CBD area showed a decrease in the survival rate in the future, and the commercial districts around university districts and IT industry clusters were divided into commercial districts with an increased survival rate in the future. The results of this study are expected to be used as quantitative evidence for more direct and realistic policy establishment.

PMID:40638714 | DOI:10.1371/journal.pone.0326307

Science. 2025 Jul 10:eadv9817. doi: 10.1126/science.adv9817. Online ahead of print.

ABSTRACT

Following the sequence and structure revolutions, predicting functionally relevant protein structure changes at scale remains an outstanding challenge. We introduce BioEmu, a deep learning system that emulates protein equilibrium ensembles by generating thousands of statistically independent structures per hour on a single GPU. BioEmu integrates over 200 milliseconds of molecular dynamics (MD) simulations, static structures and experimental protein stabilities using novel training algorithms. It captures diverse functional motions-including cryptic pocket formation, local unfolding, and domain rearrangements-and predicts relative free energies with 1 kcal/mol accuracy compared to millisecond-scale MD and experimental data. BioEmu provides mechanistic insights by jointly modelling structural ensembles and thermodynamic properties. This approach amortizes the cost of MD and experimental data generation, demonstrating a scalable path toward understanding and designing protein function.

PMID:40638710 | DOI:10.1126/science.adv9817

Br J Radiol. 2025 Jul 10:tqaf150. doi: 10.1093/bjr/tqaf150. Online ahead of print.

ABSTRACT

OBJECTIVES: To assess the diagnostic yield of thoraco-abdominal-pelvic CT (TAP-CT) in suspected infection of unclear origin in the Emergency Department (ED) and identify predictive factors for normal TAP-CT to optimize its use.

METHODS: We retrospectively categorized 517 TAP-CT studies of adult patients with non-severe infection of unclear origin based on the presence of an infectious focus or significant findings, such as neoplasia or thrombosis. Descriptive analysis, correspondence assessment between CT results and final diagnosis, and statistical modeling were performed to identify predictors of normal TAP-CT.

RESULTS: An infectious focus was identified in 55% TAP-CT scans, mainly pulmonary (46%), bilio-digestive (25%), and genitourinary (23%). Significant noninfectious findings were detected in 20%, including thrombosis (7%) and neoplasia (12%). TAP-CT showed a sensitivity of 73%, specificity of 88%, PPV of 94%, and NPV of 57%, with moderate agreement (Kappa= 0.53) between TAP-CT findings and final diagnosis. Overall, 67% of patients had an identifiable cause for infection-like symptoms. Although C-reactive protein <128 mg/L was associated with normal TAP-CT, no model reliably predicted a normal scan.

CONCLUSIONS: TAP-CT identified a relevant finding in over two-thirds of cases, reinforcing its role in diagnosing both infectious and mimicking conditions. No specific criteria could safely exclude TAP-CT, making it a valuable tool for managing patients with suspected infections of unclear origin.

ADVANCES IN KNOWLEDGE: This study is the first to assess TAP-CT’s value in suspected non-severe infections of unclear origin in the ED, highlighting its role in detecting infectious and noninfectious conditions and optimizing diagnostic strategies.

PMID:40638233 | DOI:10.1093/bjr/tqaf150

Comput Inform Nurs. 2025 Jul 9. doi: 10.1097/CIN.0000000000001358. Online ahead of print.

ABSTRACT

This study aimed to examine the effect of the course on technology use in nursing on students’ readiness for self-directed learning and attitudes toward technology. This was a quasi-experimental study. The study involved 109 first-year nursing students assigned to the intervention group (n = 53) and the control group (n = 56). Whereas the intervention group participated in the course on technology use in nursing, the control group participated in health assessment course. Data were collected with the Student Information Form, the Readiness of Self-directed Learning Scale, and Technology Attitudes Survey between March and May 2024. There was a statistically significant difference between the self-directed learning readiness and attitudes toward technology scores of the intervention and control groups (P < .05). It was found that both the self-directed learning readiness and positive attitudes toward technology scores of the students in the intervention group were significantly higher than the control group (P < .05). The study’s results indicated that course on technology use in nursing improved students’ self-directed learning readiness and positive attitudes toward technology. The integration of technology-based interventions into nursing curriculum is recommended.

PMID:40638231 | DOI:10.1097/CIN.0000000000001358

Interdiscip Cardiovasc Thorac Surg. 2025 Jul 10:ivaf133. doi: 10.1093/icvts/ivaf133. Online ahead of print.

ABSTRACT

OBJECTIVES: Current guidelines recommend concomitant surgical ablation at the time of mitral surgery for patients with atrial fibrillation, however there is a paucity of data on long-term outcomes in this population. We sought to assess long-term clinical outcomes in patients undergoing concomitant surgical ablation at the time of mitral surgery.

METHODS: The United States Centers for Medicare and Medicaid data was used to identify patients undergoing mitral repair or replacement from 2015 to 2019. After excluding prior cardiac surgery, endocarditis, and emergencies, we identified 11,410 patients undergoing isolated mitral repairs or replacement with pre-operative atrial fibrillation. Of these, 3,268(29%) received surgical ablation and 8,142(71%) did not. Propensity score matching was performed on 27 baseline characteristics. The primary outcome was freedom from death or stroke at four years. The secondary outcome was all-cause mortality at four years. Both were assessed using Cox-proportional hazard models.

RESULTS: Propensity matching yielded 3,268 well-matched patient pairs (mean age: 74, 53% female, median CHA2DS2-Vasc score 4). There was no difference in all-cause mortality at 30-days (2.7% with concomitant ablation versus 2.8% without, p = 0.762). Patients undergoing concomitant ablation at the time of surgery had significantly higher freedom from death or stroke at four years (81% vs 77%, HR: 0.84, 95% CI 0.74-0.96). However, overall freedom from death between groups was not statistically significant (84% with concomitant ablation vs 82% without, HR: 1.14, 95% CI 0.76-1.01).

CONCLUSIONS: Surgical ablation at the time of isolated mitral surgery is underutilized but associated with improved long-term outcomes.

PMID:40638230 | DOI:10.1093/icvts/ivaf133

Comput Inform Nurs. 2025 Jul 2. doi: 10.1097/CIN.0000000000001333. Online ahead of print.

ABSTRACT

There is a lack of evidence-based information on the use of technology in first-aid education. For this reason, this study aimed to describe the learning outcomes of three-dimensional technology among practical nursing students in first-aid courses. In this quasi-experimental study, first-year practical nursing students (n = 59) were divided into intervention group (n = 32) and control group (n = 27). For the intervention group, the first-aid course (a total of 16 hours per group) included three-dimensional images, three-dimensional environments, and three-dimensional printing. For the control group, the teaching was implemented using traditional methods. The data of knowledge (pre, post, and follow-up) and skills (post) were collected. The intervention group obtained statistically significantly higher scores in knowledge in follow-up test than the control group (P = .048). They also performed better on the entire resuscitation protocol (P = .0193) and in the following parts of resuscitation: student call for help, student opens the airway correctly, student checks the breathing correctly, and student has a correct depth in the chest compressions. As a conclusion, three-dimensional technology can enhance students’ first-aid knowledge and improve cardiopulmonary resuscitation skills in practical nursing education.

PMID:40638223 | DOI:10.1097/CIN.0000000000001333

Comput Inform Nurs. 2025 Jul 9. doi: 10.1097/CIN.0000000000001355. Online ahead of print.

ABSTRACT

The Internet provides many populations access to virtual health services, social resources, and health education. It is not understood how lack of household Internet adoption impacts the risk of negative health outcomes. Our study aimed to examine the relationship between Internet adoption and self-reported diabetes prevalence in US counties, while controlling for social determinants of health. This cross-sectional, retrospective study used US national county-level data obtained from the 2021 American Communities Survey and the 2021 Behavioral Risk Factors Surveillance System. Analysis included descriptive statistics, two-stage linear regression, and machine learning. A total of 3076 counties were analyzed. The results show that, in 2021, Internet adoption had a significant inverse relationship (β = -.20, P < .001) with diabetes prevalence in US counties while controlling for other social determinants of health. The results suggest that as household Internet adoption rates increase, diabetes prevalence decreases, at the county level. The relationships between social characteristics, Internet adoption, and health behaviors on diabetes prevalence in US counties warrant future research including individual-level validation and integration of health behaviors related to diabetes risk. These findings provide evidence of Internet adoption as a social determinant of health.

PMID:40638222 | DOI:10.1097/CIN.0000000000001355

Comput Inform Nurs. 2025 Jul 3. doi: 10.1097/CIN.0000000000001348. Online ahead of print.

ABSTRACT

This study investigated clinically significant association rules within pressure injury (PI) data from adult hospitalized patients to inform evidence-based prevention and management strategies. A retrospective cohort analysis was performed on a multicenter sample comprising 2386 PI cases (January 2018 to October 2023) across two tertiary hospital districts in Zhejiang Province, China. The analytical framework incorporated five patient-level demographic/clinical variables and six PI-specific characteristics. Association rule mining was conducted using the Apriori algorithm (minimum support = 10%, confidence threshold = 80%, lift >1), yielding 579 preliminary rules. Subsequent validation via χ2 testing retained 540 statistically significant associations (P < .05), of which 11 clinically actionable rules were established through Delphi consensus by a multidisciplinary expert panel. The results corroborate existing epidemiological evidence: advanced age (≥65 years), hypoalbuminemia (<35 g/L), and comorbid respiratory/neurological disorders constitute predominant risk factors for PI development. This study demonstrates the methodological rigor of association rule mining in identifying high-risk patient profiles, facilitating targeted early interventions to reduce PI incidence in inpatient populations.

PMID:40638211 | DOI:10.1097/CIN.0000000000001348