{"title":"Association of serum lysophosphatidylcholine acyltransferase 3 levels with metabolic variables and risk of type 2 diabetes mellitus: A cross-sectional study.","authors":"Haifeng Zhu, Ziyi Zhong, Jing Jin, Wei Liu, Yuan Cao, Yawen Guo, Gaonian Zhao, Qian Li","doi":"10.1371/journal.pone.0329301","DOIUrl":"https://doi.org/10.1371/journal.pone.0329301","url":null,"abstract":"<p><strong>Aims: </strong>This study aimed to explore the role of serum lysophosphatidylcholine acyltransferase 3 (LPCAT3) in glucose and lipid metabolism and its association with type 2 diabetes mellitus (T2DM).</p><p><strong>Methods: </strong>Between July and December 2024, we recruited 256 newly diagnosed T2DM patients and 252 gender- and age-matched individuals with normal glucose tolerance (NGT). Serum LPCAT3 levels were measured using ELISA. Group comparisons were conducted via t-tests or Mann-Whitney U tests. Spearman correlation analysis assessed the relationship between LPCAT3 and metabolic variables. Linear regression identified independent predictors of LPCAT3 levels. Partial Least Squares (PLS) analysis evaluated the correlations between serum LPCAT3 and obesity-related anthropometric indicators, blood glucose and lipid indicators. Logistic regression evaluated the association between LPCAT3 levels and T2DM risk, and ROC analysis determined its predictive value.</p><p><strong>Results: </strong>Median LPCAT3 level was lower in T2DM patients (21.51 ng/ml, IQR: 8.47-35.63) compared to the NGT group (24.43 ng/ml, IQR: 14.41-49.37). In NGT individuals, LPCAT3 negatively correlated with high-density lipoprotein cholesterol (HDL), fasting blood glucose (FBG), and glycated hemoglobin (HbA1c). In T2DM patients, LPCAT3 negatively correlated with body mass index (BMI) and waist circumference (WC). Linear regression identified BMI, HDL, and FBG as negative predictors of LPCAT3. PLS analysis revealed negative correlations between LPCAT3 and BMI, WC, HDL and FBG, but with large standard errors. When stratified by LPCAT3 tertiles, the lowest tertile initially showed a higher T2DM incidence than the highest tertile. However, after adjusting for obesity-related indicators, no significant difference was found between them. ROC analysis yielded an AUC of 0.580 for LPCAT3.</p><p><strong>Conclusion: </strong>Although serum LPCAT3 levels are lower in T2DM patients, its predictive capacity for T2DM is constrained. Moreover, the association between LPCAT3 and T2DM risk is likely confounded by obesity-related factors. While LPCAT3 tends to negatively correlate with BMI, HDL, and FBG, these correlations are complex and unstable.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329301"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0329271
Jia Yi Chen, Ying Su, Katsuo Tamaoka
{"title":"Positioning of Chinese time nouns and adverbs: Evidence from corpus, acceptability, and processing studies.","authors":"Jia Yi Chen, Ying Su, Katsuo Tamaoka","doi":"10.1371/journal.pone.0329271","DOIUrl":"https://doi.org/10.1371/journal.pone.0329271","url":null,"abstract":"<p><p>This study examines the syntactic placement and cognitive processing of time nouns and time adverbs in Mandarin Chinese, a language without overt tense morphology, highlighting how these temporal expressions interface with Chinese grammar. Study 1 analyzed a large-scale natural language corpus (BLCU Chinese Corpus) to determine the typical positions of time nouns and time adverbs relative to the subject. The results revealed distinct distributional patterns: time nouns occurred flexibly either before or after the subject, while time adverbs appeared predominantly in post-subject (sentence-internal) positions. Study 2 investigated native Mandarin speakers' acceptability judgments of sentences with time expressions in various positions. Sentences in which time nouns followed the subject were rated as more acceptable and supported the canonical word order, whereas pre-subject time nouns were acceptable mainly in topicalized contexts. In contrast, time adverbs were strongly preferred in post-subject positions, with only a few exceptions where certain adverbs could be fronted. Study 3 examined the real-time comprehension of these structures using reaction time and accuracy. Results showed that sentences with time expressions in non-canonical positions incurred greater processing costs, while canonical post-subject placements facilitated faster and more accurate processing. These findings suggest that the human sentence processor is sensitive to structural preferences for temporal expressions, mirroring patterns in natural use and grammatical acceptability. By integrating corpus analysis, acceptability judgments, and psycholinguistic data, this study provides a comprehensive account of how time nouns and time adverbs are positioned and processed in Chinese, offering broader implications for understanding temporal reference in tenseless languages.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329271"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754052","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0329311
{"title":"Expression of Concern: Magnitude and associated factors of disrespect and abusive care among laboring mothers at public health facilities in Borena District, South Wollo, Ethiopia.","authors":"","doi":"10.1371/journal.pone.0329311","DOIUrl":"https://doi.org/10.1371/journal.pone.0329311","url":null,"abstract":"","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329311"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754069","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0327347
Li Zhang
{"title":"Features extraction based on Naive Bayes algorithm and TF-IDF for news classification.","authors":"Li Zhang","doi":"10.1371/journal.pone.0327347","DOIUrl":"https://doi.org/10.1371/journal.pone.0327347","url":null,"abstract":"<p><p>The rapid proliferation of online news demands robust automated classification systems to enhance information organization and personalized recommendation. Although traditional methods like TF-IDF with Naive Bayes provide foundational solutions, their limitations in capturing semantic nuances and handling real-time demands hinder practical applications. This study proposes a hybrid news classification framework that integrates classical machine learning with modern advances in NLP to address these challenges. Our methodology introduces three key innovations: (1) Domain-Specific Feature Engineering, combining tailored n-grams and entity-aware TF-IDF weighting to amplify discriminative terms; (2) BERT-Guided Feature Selection, leveraging distilled BERT to identify contextually important words and resolve rare-term ambiguities; and (3) Computationally Efficient Deployment, achieving 95.2% of the accuracy of BERT at 1/52.4th of the inference cost. Evaluated on a balanced corpus of Sina News articles in 11 categories, the system demonstrates a test precision of 95.12% (vs. 84.43% for SVM+TF-IDF baseline), with statistically significant improvements confirmed by 5-fold cross-validation(p < 0.01). The critical findings reveal strong performance in distinguishing semantically distinct categories, while exposing challenges in fine-grained differentiation. The efficiency of the framework (2.1 inference latency) and scalability (linear utilization of CPU resources) validate its practicality for real-world deployment. This work bridges the gap between traditional feature engineering and transformer-based models, offering a cost-effective solution for news platforms. Future research will explore hierarchical classification and the adaptation of dynamic topics to further refine semantic boundaries.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0327347"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0329280
Gabriel Costa Souto, Ana Clara Cassimiro Nunes, Rodrigo Alberto Vieira Browne, Yuri Alberto Freire, Paulo Felipe Ribeiro Bandeira, Eduardo Caldas Costa
{"title":"Associations between fitness, movement behaviors, and immediate post-exercise blood pressure in older adults: A network perspective.","authors":"Gabriel Costa Souto, Ana Clara Cassimiro Nunes, Rodrigo Alberto Vieira Browne, Yuri Alberto Freire, Paulo Felipe Ribeiro Bandeira, Eduardo Caldas Costa","doi":"10.1371/journal.pone.0329280","DOIUrl":"https://doi.org/10.1371/journal.pone.0329280","url":null,"abstract":"<p><p>Understanding the associated factors of immediate post-exercise blood pressure, a proxy for exercise blood pressure, is relevant for improving cardiovascular risk assessment and guiding interventions. This study investigated associations between fitness and movement behavior components with immediate post-exercise blood pressure in older adults using network analysis. In this cross-sectional study, 237 older adults (66 ± 5 years), without known cardiovascular disease were included. Blood pressure was measured before and immediately after a 3-minute moderate walking test (5 METs). Moderate-to-vigorous physical activity and sedentary time were assessed using hip-worn accelerometers for one week. Proxies for fitness were: six-minute walking test, handgrip strength, and 30-second sit-to-stand test. Network analysis, a multivariate statistical approach which captures interrelations among multiple variables within a system, was used to examine direct and indirect associations between fitness, movement behaviors, and post-exercise blood pressure. Lower handgrip strength (weight: -0.136) and 30-second sit-to-stand (weight: -0.106) performances were directly associated with greater immediate post-exercise blood pressure, indicating that reduced muscular strength is linked to greater cardiovascular load after exertion. Lower handgrip strength (weight: -0.176) and reduced physical activity (weight: -0.128) were directly associated with higher resting BP, which in turn had the strongest direct association with immediate post-exercise blood pressure (weight: 0.581), suggesting an indirect pathway through resting blood pressure. Centrality indicators suggested six-minute walking test, 30-second sit-to-stand test, and sedentary time as potential interventions targets, moderate-to-vigorous physical activity as a central hub within network, and handgrip strength as a highly connected node. Our findings suggest a complex interplay - both direct and indirect (via resting blood pressure) - between fitness, movement behaviors, and immediate post-exercise blood pressure in older adults. Increasing moderate-to-vigorous physical activity, cardiorespiratory fitness, and muscle strength, along with reducing sedentary time, may be potential targets for attenuating immediate post-exercise blood pressure in this age group.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329280"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0329364
Nir Gomel, Asaf Achiron, Nadav Shemesh, Elad Eilon, Tal Yahalomi, Dana Barequet, Nadav Levinger, Dafi Porat, Shmuel Levinger, Ami Hirsch, Nir Sorkin, Eliya Levinger
{"title":"Brimonidine tartrate 0.15% drops to reduce low post-LASIK residual myopia: A retrospective study.","authors":"Nir Gomel, Asaf Achiron, Nadav Shemesh, Elad Eilon, Tal Yahalomi, Dana Barequet, Nadav Levinger, Dafi Porat, Shmuel Levinger, Ami Hirsch, Nir Sorkin, Eliya Levinger","doi":"10.1371/journal.pone.0329364","DOIUrl":"https://doi.org/10.1371/journal.pone.0329364","url":null,"abstract":"<p><strong>Purpose: </strong>Residual myopia following Laser-assisted in situ keratomileusis (LASIK) surgery poses a significant concern, with existing literature extensively detailing the use of timolol for treatment. This study aims to assess prediction factors for brimonidine tartrate 0.15% (Alphagan-P) response on post-LASIK residual myopic refraction reduction.</p><p><strong>Methods: </strong>The study included consecutive patients who received Alphagan-P during their follow-up for post-LASIK residual myopia.</p><p><strong>Results: </strong>We included 61 patients (55% male) with a mean age of 35.18 ± 10.13 Alphagan-P Treatment started at a mean of 4.94 ± 5.64 months after surgery for residual myopia of -0.53 ± 0.71D. Comparison analysis of patients who responded (n = 32, 51.6%) to patients who did not (n = 30, 48.4%) shows that responders were older (38.1 ± 9.1 vs. 32.3 ± 10.3, p = 0.03), had higher baseline myopic Spherical equivalent (SE, -0.82 ± 0.65 vs. -0.26 ± 0.66 p < 0.01), and lower uncorrected visual acuity (Uncorrected visual acuity [UCVA], 0.14 ± 0.2 vs. -0.003 ± 0.12 p = 0.01). Multiple Binary logistic regression confirmed these predictors for response (UCVA (OR=70.6, P = .006), larger SE (OR=3.8, P = .004,) and older age (OR=1.06, P = .03)).</p><p><strong>Conclusions: </strong>Alphagan-P can reduce up to 0.5D of post-LASIK residual myopia in roughly 50% of subjects. This treatment might be recommended to fine-tune outcomes for low residual myopia.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329364"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0329391
Zita Berki, Tibor Kiss, Judit Bányai, András Cseh, Krisztina Balla, Ildikó Karsai
{"title":"Effect of drought stress during critical developmental stages on morphological and grain yield-related traits in winter barley (Hordeum vulgare L.).","authors":"Zita Berki, Tibor Kiss, Judit Bányai, András Cseh, Krisztina Balla, Ildikó Karsai","doi":"10.1371/journal.pone.0329391","DOIUrl":"https://doi.org/10.1371/journal.pone.0329391","url":null,"abstract":"<p><p>As the frequency of droughts increases, the breeding of new drought-tolerant cereal varieties may become increasingly important. However, the complex effects of drought stress on grain yield-related traits are difficult to study precisely, and relatively little information is available on how drought during flowering affects plants. Therefore, 28 winter barley cultivars were included into controlled environmental tests, where their reactions were determined to single drought stress treatment applied at heading and to combined drought stresses applied at first node appearance and then at heading. Drought stress (both single and combined) significantly reduced all of the grain-yield related traits. Notably, grain yield was reduced by 48% in the two-row varieties and by 44.24% in the six-row varieties under combined drought stress, compared to the control. Our study has also demonstrated, that the combined application of drought tolerance/susceptibility indices (DT/SIs) and BLUP-based analysis provides a reliable approach for identifying stress-tolerant genotypes. We identified two main types of drought stress tolerance: the ability of preserving grain number and weight in the main ears, in parallel of maintaining the number of reproductive tillers (more tolerant), and the ability of preserving grain number and weight in the side ears (least tolerant). Both types appeared in either treatment, but not with the same intensity. Our results may provide useful information for a better understanding of this topic, which may become even more important in the context of increasingly frequent droughts.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0329391"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0327808
Jussi Hänninen, Hanna Toiviainen, Hilla Nordquist
{"title":"Emergency dispatchers as instructors of laypersons in unplanned out-of-hospital deliveries - Interdisciplinary qualitative study.","authors":"Jussi Hänninen, Hanna Toiviainen, Hilla Nordquist","doi":"10.1371/journal.pone.0327808","DOIUrl":"https://doi.org/10.1371/journal.pone.0327808","url":null,"abstract":"<p><strong>Introduction: </strong>Unplanned out-of-hospital delivery tasks are continuously increasing and are challenging to the emergency dispatcher, but there is scant research on them, especially from the perspective of instructing the layperson in delivery. In the Finnish model of one emergency response centre authority, the expertise of the emergency dispatcher is highlighted. However, it's necessary to develop collaborative and transformative expertise and cooperation in the maternity care service system to respond to risky and unexpected childbirth events. This study adopts concepts from cultural-historical activity theory in the framework of object-oriented activity systems and negotiated knotworking. Our aim is to describe and analyse the perspectives of emergency dispatchers and laypersons when the emergency dispatcher instructs the layperson in handling an unplanned out-of-hospital delivery.</p><p><strong>Methods: </strong>This is an interdisciplinary qualitative study. The study data consists of stories written by emergency dispatchers (n = 31) and semi-structured interviews with laypersons (n = 5). The data was collected using both the method of empathy-based stories and semi-structured interviews. The data was analysed using qualitative theory-driven content analysis, data-based theory-driven thematic analysis and text mining.</p><p><strong>Results: </strong>The emergency dispatchers' and laypersons' descriptions of their actions in the examined delivery situations were structured as elements of activity systems. The thematic analysis produced two themes containing discursive characteristics of negotiated knotworking, which were (i) script innovations requiring midwifery competence and (ii) emotion work.</p><p><strong>Discussion: </strong>The object of the emergency dispatcher's actions was the physical wellbeing of the person giving birth and the newborn, while the layperson's object was the childbirth experience, including the aforementioned and shared with the person giving birth. The formal script of childbirth services does not serve negotiated knotworking. Further, a midwife's participation in an emergency call, including video consultation, is desired. The emergency dispatcher should respect the object-oriented conscious agency of the layperson, providing ad hoc information for childbirth.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0327808"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0326936
Nafiseh Jabbari Tofighi, Reda Alhajj
{"title":"Investigating the impact of social media images on users' sentiments towards sociopolitical events based on deep artificial intelligence.","authors":"Nafiseh Jabbari Tofighi, Reda Alhajj","doi":"10.1371/journal.pone.0326936","DOIUrl":"https://doi.org/10.1371/journal.pone.0326936","url":null,"abstract":"<p><p>This paper presents the findings of the research aimed at investigating the influence of visual content, posted on social media in shaping users' sentiments towards specific sociopolitical events. The study analyzed various sociopolitical topics by examining posts containing relevant hashtags and keywords, along with their associated images and comments. Using advanced machine learning and deep learning methods for sentiment analysis, textual data were classified to determine the expressed sentiments. Additionally, the correlation between posted visual content and user sentiments has been studied. A particular emphasis was placed on understanding how these visuals impact users' attitudes toward the events. The research resulted in a comprehensive dataset comprising labeled images and their comments, offering valuable insights into the dynamics of public opinion formation through social media. This study investigates the influence of social media images on user sentiment toward sociopolitical events using deep learning-based sentiment analysis. By analyzing posts from movements such as Black Lives Matter, Women's March, Climate Change Protests, and Anti-war Demonstrations, we identified a strong correlation between visual content and public sentiment. Our results reveal that Anti-war Demonstrations exhibit the highest correlation (PLCC: 0.709, SROCC: 0.723), while Climate Change Protests display the lowest alignment (PLCC: 0.531, SROCC: 0.611). Overall, the study finds a consistent positive correlation (PLCC range: 0.615-0.709, SROCC: 0.611-0.723) across movements, indicating the significant role of visual content in shaping the public opinion.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0326936"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
PLoS ONEPub Date : 2025-07-30eCollection Date: 2025-01-01DOI: 10.1371/journal.pone.0326952
Pei Zhao, Hao Song, Guang Ling
{"title":"Machine learning approaches for predicting the link of the global trade network of liquefied natural gas.","authors":"Pei Zhao, Hao Song, Guang Ling","doi":"10.1371/journal.pone.0326952","DOIUrl":"https://doi.org/10.1371/journal.pone.0326952","url":null,"abstract":"<p><p>With the rising geopolitical tensions, predicting future trade partners has become a critical topic for the global community. Liquefied natural gas (LNG), recognized as the cleanest burning hydrocarbon, plays a significant role in the transition to a cleaner energy future. As international trade in LNG becomes increasingly volatile, it is essential to assist governments in identifying potential trade partners and analyzing the trade network. Traditionally, forecasts of future mineral and energy resource trade networks have relied on similarity indicators (e.g., CN, AA). This study employs complex network theory to illustrate the characteristics of nodes and edges, as well as the evolution of global LNG trade networks from 2001 to 2020. Utilizing node and edge data from these networks, this research applies machine learning algorithms to predict future links based on local and global similarity-based indices (e.g., CN, JA, PA). The findings indicate that random forest and decision tree algorithms, when used with local similarity-based indices, demonstrate strong predictive performance. The reliability of these algorithms is validated through the Receiver Operating Characteristic Curve (ROC). Additionally, a graph attention network model is developed to predict potential links using edge and motif data. The results indicate robust predictive performance. This study demonstrates that machine learning algorithms-specifically random forest and decision tree-outperform in predicting links within the global LNG trade network based on local information proximity, while the graph attention network, a deep learning model, exhibits stable optimization and effective feature learning. These findings suggest that machine learning approaches hold significant promise for mineral trade network analysis.</p>","PeriodicalId":20189,"journal":{"name":"PLoS ONE","volume":"20 7","pages":"e0326952"},"PeriodicalIF":2.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}