Search

Abstract Predicting the blood–brain barrier (BBB) permeability of small-molecule compounds using a novel artificial intelligence platform is necessary for drug discovery. Machine learning and a large language model on artificial intelligence (AI) tools improve the accuracy and shorten the time for new drug development. The primary goal of this research is to develop artificial intelligence (AI) computing models and novel deep learning architectures capable of predicting whether molecules can permeate the human blood–brain barrier (BBB). The in silico (computational) and in vitro (experimental) results were validated by the Natural Products Research Laboratories (NPRL) at China Medical University Hospital (CMUH). The transformer-based MegaMolBART was used as the simplified molecular input line entry system (SMILES) encoder with an XGBoost classifier as an in silico method to check if a molecule could cross through the BBB. We used Morgan or Circular fingerprints to apply the Morgan algorithm to a set of atomic invariants as a baseline encoder also with an XGBoost classifier to compare the results. BBB permeability was assessed in vitro using three-dimensional (3D) human BBB spheroids (human brain microvascular endothelial cells, brain vascular pericytes, and astrocytes). Using multiple BBB databases, the results of the final in silico transformer and XGBoost model achieved an area under the receiver operating characteristic curve of 0.88 on the held-out test dataset. Temozolomide (TMZ) and 21 randomly selected BBB permeable compounds (Pred scores = 1, indicating BBB-permeable) from the NPRL penetrated human BBB spheroid cells. No evidence suggests that ferulic acid or five BBB-impermeable compounds (Pred scores

It is well-established that disrupted autonomic nervous system (ANS) reactivity exacerbates risk for long-term maladjustment following childhood adversity (CA). However, few studies have integrated measures of both the sympathetic (SNS) and parasympathetic (PNS) branches of the ANS, resulting in a unidimensional understanding of ANS functioning as a mechanism of risk. Further, past work has primarily measured CA only at the aggregate level (e.g. “total CA”), necessitating further research to accurately characterize this risk pathway. The present study examines how CA, measured cumulatively and dimensionally (i.e. CA characterized by threat versus deprivation), moderates the association between the SNS and PNS at rest and in response to acute social and nonsocial stressors. Participants included 97 adolescents ages 10-15 (Mage = 12.22, SDage = 1.68) experiencing a range of CA and one accompanying caregiver. Participants completed questionnaires assessing prior CA exposure. SNS and PNS responses were then continuously measured during rest and two stress tasks. First, results indicate a blunting effect of cumulative CA and CA characterized by threat (e.g. physical abuse) on resting SNS activity. Second, in moderation analyses assessing ANS coordination, threat exposure emerged as a significant moderator of the association between SNS and PNS reactivity to social stress. Results suggest that CA characterized by threat may specifically impact physiologic regulation by disrupting the coordination of the two branches of the ANS. Disentangling the independent and concurrent engagement of biological stress response systems following CA remains an important target for research to identify the etiology of aberrant stress reactivity patterns.

Federated learning (FL) is a method used for training artificial intelligence models with data from multiple sources while maintaining data anonymity, thus removing many barriers to data sharing. Here we used data from 20 institutes across the globe to train a FL model, called EXAM (electronic medical record (EMR) chest X-ray AI model), that predicts the future oxygen requirements of symptomatic patients with COVID-19 using inputs of vital signs, laboratory data and chest X-rays. EXAM achieved an average area under the curve (AUC) >0.92 for predicting outcomes at 24 and 72 h from the time of initial presentation to the emergency room, and it provided 16% improvement in average AUC measured across all participating sites and an average increase in generalizability of 38% when compared with models trained at a single site using that site's data. For prediction of mechanical ventilation treatment or death at 24 h at the largest independent test site, EXAM achieved a sensitivity of 0.950 and specificity of 0.882. In this study, FL facilitated rapid data science collaboration without data exchange and generated a model that generalized across heterogeneous, unharmonized datasets for prediction of clinical outcomes in patients with COVID-19, setting the stage for the broader use of FL in healthcare.

The COVID-19 pandemic has brought unprecedented levels of stress to individuals in the U.S. and throughout the world. These high stress levels place individuals at risk for symptoms of anxiety, depression, and other psychiatric disorders. The current study applies a control-based model of coping to contribute to the development of evidence-based interventions to promote resilience. Data were collected online from April 22 through July 12, 2020. Data from two samples of U. S. community adults who completed an online battery of standardized questionnaires were combined (N = 709). More than a quarter reported moderate to severe levels of depression symptoms, and more than one-fifth reported moderate to severe levels of anxiety symptoms; symptom levels were higher among adults who reported more COVID-19-related stress. As hypothesized, multiple regression analyses indicated that greater use of primary and secondary control coping was associated with lower symptom levels, whereas greater use of disengagement coping was associated with higher symptom levels, above and beyond the association of stress with symptoms. Race and ethnicity emerged as important moderators of these associations, indicating that what constitutes adaptive coping varies according to characteristics of the individual. Implications for public health policy and clinical practice are discussed., Author(s): Mary Jo Coiro [sup.1] , Kelly H. Watson [sup.2] , Abby Ciriegio [sup.2] , Maile Jones [sup.2] , Amy R. Wolfson [sup.1] , Jennie Reisman [sup.2] , Bruce E. [...]

Abstract There is an increasing interest in developing artificial intelligence (AI) systems to process and interpret electronic health records (EHRs). Natural language processing (NLP) powered by pretrained language models is the key technology for medical AI systems utilizing clinical narratives. However, there are few clinical language models, the largest of which trained in the clinical domain is comparatively small at 110 million parameters (compared with billions of parameters in the general domain). It is not clear how large clinical language models with billions of parameters can help medical AI systems utilize unstructured EHRs. In this study, we develop from scratch a large clinical language model—GatorTron—using >90 billion words of text (including >82 billion words of de-identified clinical text) and systematically evaluate it on five clinical NLP tasks including clinical concept extraction, medical relation extraction, semantic textual similarity, natural language inference (NLI), and medical question answering (MQA). We examine how (1) scaling up the number of parameters and (2) scaling up the size of the training data could benefit these NLP tasks. GatorTron models scale up the clinical language model from 110 million to 8.9 billion parameters and improve five clinical NLP tasks (e.g., 9.6% and 9.5% improvement in accuracy for NLI and MQA), which can be applied to medical AI systems to improve healthcare delivery. The GatorTron models are publicly available at: https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/models/gatortron_og .

Many interventions are characterized by repeated observations on the same individuals (e.g., baseline, mid-intervention, two to three post-intervention observations), which offer the opportunity to consider differences in how individuals vary over time. Effective interventions may not be limited to changing means, but instead may also include changes to how variables affect each other over time. Continuous time models offer the opportunity to specify differing underlying processes for how individuals change from one time to the next, such as whether it is the level or change in a variable that is related to changes in an outcome of interest. After introducing continuous time models, we show how different processes can produce different expected covariance matrices. Thus, models representing differing underlying processes can be compared, even with a relatively small number of repeated observations. A substantive example comparing models that imply different underlying continuous time processes will be fit using panel data, with parameters reflecting differences in dynamics between control and intervention groups.

Regulatory efforts are hypothesized to affect associations between emotions and physiology (i.e., concordance) to facilitate adaptive functioning. Assessing the role of coping on physiological–emotional concordance during ecologically relevant scenarios can elucidate whether concordance can serve as a biomarker of risk or resilience. The present study assessed self‐reported coping as a moderator of minute‐to‐minute associations between autonomic nervous system activity and emotions (i.e., physiological–emotional concordance) in caregivers (N = 97) and adolescents (N = 97; ages 10–15) during a dyadic conflict task. Models included physiological variables (sympathetic, skin conductance level [SCL]; and parasympathetic, respiratory sinus arrhythmia [RSA]) and their interaction (SCL × RSA) as predictors of emotions, with coping variables as moderators. Caregivers' use of primary control coping (e.g., problem solving and emotional expression) and secondary control coping (e.g., cognitive reappraisal and acceptance) use in response to family stress predicted more positive emotional experiences during the laboratory conflict task. Adolescents' use of secondary control coping moderated the SCL–emotion association, such that increases in momentary SCL were associated with more positive emotion ratings for youth reporting higher secondary control coping. For youth who report more adaptive trait‐level coping skills, momentary changes in SCL may reflect active engagement and attentiveness to facilitate more positive emotional experiences. Findings advance our understanding of the interrelationships between physiological responses and psychological experiences during relevant, interactive scenarios. Autonomic responses are differentially related to affective states depending on the coping strategies that adolescents employ, suggesting that concordance may be associated with intervention targets (i.e., coping skills). Regulatory efforts are hypothesized to affect physiological–emotional associations to facilitate adaptive functioning. Results suggest that autonomic responses differentially relate to emotions based on coping. For adolescents utilizing more secondary control coping skills, momentary skin conductance level changes may reflect active engagement and attentiveness to facilitate more positive emotional experiences. [ABSTRACT FROM AUTHOR]

Background: Huntington's disease (HD) presents patients and individuals at risk for HD with significant levels of stress. However, relatively little research has examined how individuals cope with stress related to the disease or the association of specific coping strategies with psychological symptoms. Objective: This study examined the ways in which HD patients and at-risk individuals cope with HD-related stress using a control-based model of coping and the association of coping strategies with symptoms of depression and anxiety. Methods: HD patients (n = 49) and at-risk individuals (n = 76) completed the Responses to Stress Questionnaire – Huntington's Disease Version to assess coping strategies in response to HD-related stress, as well as standardized measures of depression and anxiety symptoms. Patient health records were accessed to obtain information related to disease characteristics. Results: Patients and at-risk individuals reported using comparable levels of primary control coping, secondary control coping, and disengagement coping strategies. In linear regression analyses, only secondary control coping was significantly associated with lower depression (β= –0.62, p < 0.001) and anxiety (β= –0.59, p < 0.001) symptoms in patients and at-risk individuals (β= –0.55, p < 0.001 and β= –0.50, p < 0.001, respectively). Conclusions: Secondary control coping may be beneficial for both HD patients and at-risk individuals. Future research using the control-based model of coping in longitudinal studies with the HD population is needed, and future interventions could test the effects of cognitive reframing and acceptance as coping strategies for families affected by HD. [ABSTRACT FROM AUTHOR]

The objective was to characterize the relation between different sources of school-based social support (friends, peers, and teachers) and bereaved siblings' grief and grief-related growth and to examine whether nonparental sources of social support buffer the effects of low parent support on bereaved siblings. Families (N = 85) were recruited from cancer registries at 3 pediatric institutions 3-12 months after a child's death. Bereaved siblings were 8-18 years old (M = 12.39, SD = 2.65) and majority female (58%) and White (74%). During home visits, siblings reported their perceptions of social support from parental and nonparental sources using the Social Support Scale for Children, as well as grief and grief-related growth using the Hogan Sibling Inventory of Bereavement. Parent, friend, and teacher support were positively correlated with grief-related growth, whereas parent and peer support were negatively correlated with grief for adolescents. Teacher and friend support significantly moderated the association between parent support and grief such that teacher and friend support accentuated the positive effects of parent support. Friend and peer support moderated associations between parent support and grief/growth for adolescents but not children. School-based social support, namely from friends, peers, and teachers, appears to facilitate the adjustment of bereaved siblings. Findings suggest that bereaved siblings may benefit from enhanced support from teachers and friends regardless of age, with middle/high school students particularly benefitting from increased support from close friends and peers.

Background: Huntington's disease (HD) is a neurodegenerative disease that presents families with significant numbers of stressful events. However, relatively little empirical research has characterized the stressors encountered by members of HD-affected families and their correlations with psychological symptoms. Objective: This study examined frequencies of specific stressors in HD patients and at-risk individuals and the correlates of these stressors with demographics, disease characteristics, and symptoms of depression and anxiety. Methods: HD patients (n = 57) and at-risk individuals (n = 81) completed the Responses to Stress Questionnaire –Huntington's Disease Version to assess HD-related stressors. Participants completed measures of depression and anxiety symptoms. Patient health records were accessed to obtain information related to disease characteristics. Results: Patients endorsed a mean number of 5.05 stressors (SD = 2.74) out of the 10-item list. Demographics were not related to total stressors, but disease characteristics were significantly related to specific stressors. At-risk individuals endorsed a mean number of 3.20 stressors (SD = 2.65) out of the 11-item list. Age and sex were significantly related to specific stressors. Total number of stressors was significantly related to depression (β=0.67, p < 0.001) and anxiety symptoms (β=0.58, p < 0.001) in patients and at-risk individuals (β=0.35, p = 0.003 and β=0.32, p = 0.006, respectively). Conclusions: hese findings emphasize the significant burden of stress experienced by HD patients and at-risk individuals. We highlight a need for more specific stress-based measures and psychosocial support interventions for HD-affected families. [ABSTRACT FROM AUTHOR]