Health data science最新文献

Background: In real-world drug discovery, human experts typically grasp molecular knowledge of drugs and proteins from multimodal sources including molecular structures, structured knowledge from knowledge bases, and unstructured knowledge from biomedical literature. Existing multimodal approaches in AI drug discovery integrate either structured or unstructured knowledge independently, which compromises the holistic understanding of biomolecules. Besides, they fail to address the missing modality problem, where multimodal information is missing for novel drugs and proteins. Methods: In this work, we present KEDD, a unified, end-to-end deep learning framework that jointly incorporates both structured and unstructured knowledge for vast AI drug discovery tasks. The framework first incorporates independent representation learning models to extract the underlying characteristics from each modality. Then, it applies a feature fusion technique to calculate the prediction results. To mitigate the missing modality problem, we leverage sparse attention and a modality masking technique to reconstruct the missing features based on top relevant molecules. Results: Benefiting from structured and unstructured knowledge, our framework achieves a deeper understanding of biomolecules. KEDD outperforms state-of-the-art models by an average of 5.2% on drug-target interaction prediction, 2.6% on drug property prediction, 1.2% on drug-drug interaction prediction, and 4.1% on protein-protein interaction prediction. Through qualitative analysis, we reveal KEDD's promising potential in assisting real-world applications. Conclusions: By incorporating biomolecular expertise from multimodal knowledge, KEDD bears promise in accelerating drug discovery.

Background: Digital exclusion is a global issue that disproportionately affects older individuals especially in low- and middle-income nations. However, there is a wide gap in current research regarding the impact of digital exclusion on the mental health of older adults in both high-income and low- and middle-income countries. Methods: We analyzed data from 5 longitudinal cohorts: the Health and Retirement Study (HRS), the English Longitudinal Study of Aging (ELSA), the Survey of Health, Ageing and Retirement in Europe (SHARE), the China Health and Retirement Longitudinal Study (CHARLS), and the Mexican Health and Aging Study (MHAS). These cohorts consisted of nationwide samples from 24 countries. Digital exclusion was defined as the self-reported lack of access to the internet. Depressive symptoms were assessed using comparable scales across all cohorts. We used generalized estimating equation models, fitting a Poisson model, to investigate the association between the digital exclusion and depressive symptoms. We adjusted for the causal directed acyclic graph (DAG) minimal sufficient adjustment set (MSAS), which includes gender, age, retirement status, education, household wealth, social activities, and weekly contact with their children. Results: During the study period (2010-2018), 122,242 participants underwent up to 5 rounds of follow-up. Digital exclusion varied greatly across countries, ranging from 21.1% in Denmark to 96.9% in China. The crude model revealed a significant association between digital exclusion and depressive symptoms. This association remained statistically significant in the MSAS-adjusted model across all cohorts: HRS [incidence rate ratio (IRR), 1.37; 95% confidence interval (CI), 1.28 to 1.47], ELSA (IRR, 1.32; 95% CI, 1.23 to 1.41), SHARE (IRR, 1.30; 95% CI, 1.27 to 1.33), CHARLS (IRR, 1.62; 95% CI, 1.38 to 1.91), and MHAS (IRR, 1.31; 95% CI, 1.26 to 1.37); all Ps < 0.001. Notably, this association was consistently stronger in individuals living in lower wealth quintile households across all 5 cohorts and among those who do not regularly interact with their children, except for ELSA. Conclusions: Digital exclusion is globally widespread among older adults. Older individuals who are digitally excluded are at a higher risk of developing depressive symptoms, particularly those with limited communication with their offspring and individuals living in lower wealth quintile households. Prioritizing the provision of internet access to older populations may help reduce the risks of depression symptoms, especially among vulnerable groups with limited familial support and with lower income.

Background: Although loneliness and social isolation are proposed as important risk factors for metabolic diseases, their associations with the risk of non-alcoholic fatty liver disease (NAFLD) have not been elucidated. The aims of this study were to determine whether loneliness and social isolation are independently associated with the risk of NAFLD and to explore potential mediators for the observed associations. Methods: In this large prospective cohort analysis with 405,073 participants of the UK Biobank, the status of loneliness and social isolation was assessed through self-administrated questionnaires at study recruitment. The primary endpoint of interest was incident NAFLD. Multivariable-adjusted Cox proportional hazard regression models were used to calculate hazard ratios (HRs) and 95% confidence intervals for the associations between loneliness, social isolation, and risk of NAFLD. Results: During a median follow-up of 13.6 years, there were 5,570 cases of NAFLD identified. In the multivariable-adjusted model, loneliness and social isolation were both statistically significantly associated with an increased risk of NAFLD (HR = 1.22 and 1.13, respectively). No significant multiplicative or additive interaction was found between loneliness and social isolation on the risk of NAFLD. The mediation analysis estimated that 30.4%, 16.2%, 5.3%, 4.1%, 10.5%, and 33.2% of the loneliness-NAFLD association was mediated by unhealthy lifestyle score, obesity, current smoking, irregular physical activity, suboptimal sleep duration, and depression, respectively. On the other hand, 25.6%, 10.1%, 15.5%, 10.1%, 8.1%, 11.6%, 9.6%, 4.8%, and 3.0% of the social isolation-NAFLD association was mediated by unhealthy lifestyle score, obesity, current smoking, irregular physical activity, suboptimal sleep duration, depression, C-reactive protein, count of white blood cells, and count of neutrophils, respectively. Conclusions: Our study demonstrated that loneliness and social isolation were associated with an elevated risk of NAFLD, independent of other important risk factors. These associations were partially mediated by lifestyle, depression, and inflammatory factors. Our findings substantiate the importance of loneliness and social isolation in the development of NAFLD.

[This corrects the article DOI: 10.34133/hds.0099.].