The online version's supplementary materials are located at 101007/s12144-023-04353-2.
The COVID-19 pandemic's imposition of online learning created new challenges to the safety and well-being of young people, who spent considerable time online, with cyberbullying emerging as a major concern for parents, teachers, and the students themselves. Two online studies in Portugal investigated the frequency, contributing factors, and repercussions of cyberbullying during the COVID-19 lockdowns. In-depth analysis of Study 1's findings is needed to derive meaningful conclusions.
A 2020 study, examining the prevalence of cyberbullying among adolescents during the initial lockdown, investigated predictors, psychological distress symptoms, and potential mitigating factors related to cyberbullying. Regarding Study 2, the output should be a list of sentences in JSON format.
Examining the second lockdown period of 2021, the research investigated the prevalence of cyberbullying, associated risk factors, and indications of psychological distress. Cyberbullying was frequently observed in the study's results; participants who were victims of cyberbullying exhibited more pronounced symptoms of lockdown-induced psychological distress, including sadness and loneliness; surprisingly, those who experienced cyberbullying while simultaneously receiving high levels of parental and social support displayed lower psychological distress, including thoughts of suicide. These results provide new insight into online bullying among young people during the COVID-19 lockdowns, augmenting previous studies.
The online version includes additional materials at 101007/s12144-023-04394-7.
For the online version, supplementary materials are provided at the link 101007/s12144-023-04394-7.
A key characteristic of posttraumatic stress disorder (PTSD) is the disruption of cognitive processes. To probe the relationship between military-related PTSD and cognitive functions such as visual working memory and visual imagery, two studies were performed. Participants, being military personnel, documented their PTSD diagnosis history and subsequently completed the self-administered PTSD screening tool, the PTSD Checklist – Military Version. Study 1 included 138 personnel who additionally performed a memory span task and a 2-back task using colored words. Stroop interference was implemented via the semantic content of these words. Within Study 2, a distinct group of 211 personnel completed measurements of perceived imagery vividness and the spontaneous utilization of visual imagery for their tasks. Replication of interference effects on working memory was absent in the PTSD-diagnosed military personnel group. Analysis via ANCOVA and structural equation modeling indicated that PTSD-related intrusions negatively influenced working memory capacity, whereas PTSD arousal exhibited a correlation with spontaneous visual imagery. We interpret these results to mean that intrusive flashbacks disrupt working memory not through limits on capacity or by interfering with functions like inhibition, but through the introduction of distracting task-unrelated memories and emotions. Visual imagery, seemingly independent of these flashbacks, may coexist with PTSD arousal symptoms, which could take the form of flashforwards depicting feared or anticipated threats.
The integrative parenting model has underscored the pivotal importance of parental engagement (measured by quantity) and parenting approaches (characterized by quality) on adolescent psychological well-being. This study's initial focus was on utilizing a person-centered approach to determine categories of parental involvement (quantitatively) and parenting styles (qualitatively). The study's second aspect was a deep dive into the relationship between diverse parenting styles and how adolescents fared psychologically. A cross-sectional online survey, conducted in mainland China, targeted families (N=930), comprising fathers, mothers, and adolescents (50% female, mean age = 14.37231). The parental involvement levels of mothers and fathers were reported; adolescents evaluated both parents' parenting styles and their own levels of anxiety, depression, and loneliness. Utilizing standardized scores of paternal and maternal involvement and styles (warmth and rejection), latent profile analysis was employed to delineate parenting profiles. Genetic affinity A regression mixture model was used to scrutinize the interrelationships between varying parenting profiles and adolescent psychological adaptations. The parenting behaviors could be grouped into four classes: warm involvement (526%), neglecting non-involvement (214%), rejecting non-involvement (214%), and rejecting involvement (46%). Adolescents categorized in the warm involvement group displayed the least anxiety, depression, and loneliness. Among adolescents, those who rejected involvement in the group scored the highest on measures of psychological adjustment. Adolescents experiencing neglect in the context of non-involvement reported lower anxiety symptom levels than those facing rejection in the context of non-involvement. medication safety Adolescents receiving warm involvement displayed the most favorable adjustment outcomes, whereas those in the rejecting involvement group exhibited the least favorable adjustment outcomes. Adolescent mental health improvement initiatives necessitate a dual focus on parental engagement and the specific parenting styles utilized.
For improved comprehension and prediction of disease progression, specifically regarding the lethal cancer with its high mortality rate, incorporating multi-omics data, with its comprehensive disease signals, is essential. While recent approaches are available, they unfortunately do not effectively integrate multi-omics data to provide accurate cancer survival predictions, thereby lowering the precision and efficacy of survival estimations derived from such data.
A deep learning model, incorporating multimodal representation and integration techniques, was constructed in this work to anticipate the survival of patients using multi-omics data. Initially, we constructed an unsupervised learning module to derive high-level feature representations from omics data across various modalities. The unsupervised learning phase produced feature representations, which were then combined into a single compact vector using an attention-based method. Finally, this vector was inputted into fully connected layers for survival prediction. The model, trained using a multimodal approach, accurately predicted pancancer survival rates, outperforming those models trained on single data modalities. Our method, compared to leading-edge methods via concordance index and 5-fold cross-validation, achieved superior performance on most cancer types in our testing datasets.
In the realm of survival prediction, ZhangqiJiang07's GitHub project, MultimodalSurvivalPrediction, meticulously investigates the use of various data types in forecasting outcomes.
Additional information regarding this topic is provided in the supplementary data.
online.
Online, supplementary data are accessible at the Bioinformatics resource.
Spatially resolved transcriptomics (SRT) technologies, a burgeoning area, effectively measure gene expression profiles, while precisely retaining tissue spatial localization information, often from multiple tissue sections. We have previously created SC.MEB, an empirical Bayes methodology applied to SRT data analysis, employing a hidden Markov random field structure. iSC.MEB, an extension to SC.MEB integrating hidden Markov random fields and empirical Bayes, permits simultaneous spatial clustering and batch effect estimation from low-dimensional representations of multiple SRT datasets for user benefit. Two SRT datasets are used to illustrate iSC.MEB's capability in accurately identifying cell/domain structures.
An open-source R package hosts the iSC.MEB implementation, with its source code freely downloadable from https//github.com/XiaoZhangryy/iSC.MEB. To access the documentation and illustrative examples (vignettes) for our package, please visit https://xiaozhangryy.github.io/iSC.MEB/index.html.
Data supplementary to this document is available at
online.
The supplementary data are located online at Bioinformatics Advances.
The revolutionary progress seen in natural language processing (NLP) is largely due to the achievements of transformer-based language models, including the vanilla transformer, BERT, and GPT-3. The impressive interpretability and adaptability of these models, stemming from inherent similarities between biological sequences and natural languages, have resulted in a new wave of their application within bioinformatics research. To offer a timely and comprehensive assessment, we present key progressions in transformer-based language models. This includes a thorough explanation of the transformer's structure and a synopsis of their substantial impact across bioinformatics research, encompassing tasks from basic sequence analysis to innovative drug discovery techniques. O-Propargyl-Puromycin clinical trial The breadth and depth of transformer applications in bioinformatics, while substantial, present consistent hurdles, including the heterogeneity of training data, the substantial computational burden, and the limitations in model interpretability, offering opportunities for further research. The broader community of NLP researchers, bioinformaticians, and biologists is hoped to be brought together to advance future research and development in transformer-based language models, prompting the creation of novel bioinformatics applications unavailable through traditional approaches.
Supplementary data are available at the referenced website.
online.
Online, Bioinformatics Advances provides access to the supplementary data.
Part 1 of Report 4 explores the development and refinement of causal criteria, referencing the influential framework established by A.B. Hill in 1965. The widely acknowledged textbook for modern epidemiology, B. MacMahon et al. (1970-1996), while frequently cited in the context of this topic, was scrutinized regarding its criteria, and it was determined that no novel approaches were offered. The criteria proposed by M. Susser, encompassing three fundamental points—association (or probability of causality), temporal precedence, and directionality of effect—demonstrate a degree of simplicity, while two supplementary criteria, pivotal to the advancement of Popperian epidemiology, namely the hypothesis's resilience under diverse testing methodologies (a refinement incorporated into Hill's criterion of consistency) and its predictive power, showcase a more theoretical underpinning and practical limitations in epidemiology and public health applications.