Temporal Protein Complex Identification Based on Dynamic Heterogeneous Protein Information Network Representation Learning.

Protein complexes, as the fundamental units of cellular function and regulation, play a crucial role in understanding the normal physiological functions of cells. Existing methods for protein complex identification attempt to introduce other biological information on top of the protein-protein interaction (PPI) network to assist in evaluating the degree of association between proteins. However, these methods usually treat protein interaction networks as flat homogeneous static networks. They cannot distinguish the roles and importance of different types of biological information, nor can they reflect the dynamic changes of protein complexes. In recent years, heterogeneous network representation learning has achieved great success in processing complex heterogeneous information and mining deep semantics. We thus propose a temporal protein complex identification method based on Dynamic Heterogeneous Protein information network Representation Learning, DHPRL. DHPRL naturally integrates multiple types of heterogeneous biological information in the cellular temporal dimension. It simultaneously models the temporal dynamic properties of proteins and the heterogeneity of biological information to improve the understanding of protein interactions and the accuracy of complex prediction. Firstly, we construct Dynamic Heterogeneous Protein Information Network (DHPIN) by integrating temporal gene expression information and GO attribute information. Then we design a dual-view collaborative contrast mechanism. Specifically, proposing to learn protein representations from two views of DHPIN (1-hop relation view and meta-path view) to model the consistency and specificity between nearest-neighbour bio information and deeper biological semantics. The dynamic PPI network is thereafter re-weighted based on the learned protein representations. Finally, we perform protein identification on the re-weighted dynamic PPI network. Extensive experimental results demonstrate that DHPRL can effectively model complicated biological information and achieve state-of-the-art performance in most cases. The source code and datasets for DHPR are available at https://github.com/LI-jasm/DHPRL.

Protein Complex Identification Based on Heterogeneous Protein Information Network.

Protein complexes are the foundation of all cellular activities, and accurately identifying them is crucial for studying cellular systems. The efficient discovery of protein complexes is a focus of research in the field of bioinformatics. Most existing methods for protein complex identification are based on the structure of the protein-protein interaction (PPI) network, whereas some methods attempt to integrate biological information to enhance the features of the protein network for complex identification. Existing protein complex identification methods are unable to fully integrate network topology information and biological attribute information. Most of these methods are based on homogeneous networks and cannot distinguish the importance of different attributes and protein nodes. To address these issues, a GO attribute Heterogeneous Attention network Embedding (GHAE) method based on heterogeneous protein information networks is proposed. First, GHAE incorporates Gene Ontology (GO) information into the PPI network, constructing a heterogeneous protein information network. Then, GHAE uses a dual attention mechanism and heterogeneous graph convolutional representation learning method to learn protein features and to identify protein complexes. The experimental results show that building heterogeneous protein information networks can fully integrate valuable biological information. The heterogeneous graph embedding learning method can simultaneously mine the features of protein and GO attributes, thereby improving the performance of protein complex identification.

Residential greenness mitigates mortality risk from short-term airborne particulate exposure: An individual-level case-crossover study.

BACKGROUND: Studies suggested that greenness could reduce death risks related to ambient exposure to particulate matter (PM), while the available evidence was mixed across the globe and substantially exiguous in low- and middle-income countries. By conceiving an individual-level case-crossover study in central China, this analysis primarily aimed to quantify PM-mortality associations and examined the modification effect of greenness on the relationship. METHODS: We investigated a total of 177,058 nonaccidental death cases from 12 counties in central China, 2008-2012. Daily residential exposures to PM2.5 (aerodynamic diameter <2.5 µm), PMc (aerodynamic diameter between 2.5 and 10 µm), and PM10 (aerodynamic diameter <10 µm) were assessed at a 1 × 1-km resolution through satellite-derived machine-learning models. Residential surrounding greenness was assessed using satellite-derived enhanced vegetation index (EVI) and normalized difference vegetation index (NDVI) at multiple buffer sizes (250, 500, and 1000 m). To quantify the acute mortality risks associated with short-term exposure to PM2.5, PMc, and PM10, a time-stratified case-crossover design was utilized in conjunction with a conditional logistic regression model in our main analyses. To investigate the effect modification of greenness on PM-mortality associations, we grouped death cases into low, medium, and high greenness levels using cutoffs of 25th and 75th percentiles of NDVI or EVI exposure, and examined potential effect heterogeneity in PM-related mortality risks among these groups. RESULTS: Mean concentrations (standard deviation) on the day of death were 73.8 (33.4) µg/m3 for PM2.5, 43.9 (17.3) µg/m3 for PMc, and 117.5 (44.9) µg/m3 for PM10. Size-fractional PM exposures were consistently exhibited significant associations with elevated risks of nonaccidental and circulatory mortality. For every increase of 10-µg/m3 in PM exposure, percent excess risks of nonaccidental and circulatory mortality were 0.271 (95% confidence interval [CI]: 0.010, 0.533) and 0.487 (95% CI: 0.125, 0.851) for PM2.5 at lag-01 day, 0.731 (95% CI: 0.108, 1.359) and 1.140 (95% CI: 0.267, 2.019) for PMc at lag-02 day, and 0.271 (95% CI: 0.010, 0.533) and 0.386 (95% CI: 0.111, 0.662) for PM10 at lag-01 day, respectively. Compared to participants in the low-level greenness areas, those being exposed to higher greenness were found to be at lower PM-associated risks of nonaccidental and circulatory mortality. Consistent evidence for alleviated risks in medium or high greenness group was observed in subpopulations of female and younger groups (age <75). CONCLUSIONS: Short-term exposure to particulate air pollution was associated with elevated risks of nonaccidental and circulatory death, and individuals residing in higher neighborhood greenness possessed lower risk of PM-related mortality. These findings emphasized the potential public health advantages through incorporating green spaces into urban design and planning.

Short-term effects of fine particulate matter constituents on myocardial infarction death.

Existing evidence suggested that short-term exposure to fine particulate matter (PM2.5) may increase the risk of death from myocardial infarction (MI), while PM2.5 constituents responsible for this association has not been determined. We collected 12,927 MI deaths from 32 counties in southern China during 2011-2013. County-level exposures of ambient PM2.5 and its 5 constituents (i.e., elemental carbon (EC), organic carbon (OC), sulfate (SO42-), ammonium (NH4+), and nitrate (NO3-)) were aggregated from gridded datasets predicted by Community Multiscale Air Quality Modeling System. We employed a space-time-stratified case-crossover design and conditional logistic regression models to quantify the association of MI mortality with short-term exposure to PM2.5 and its constituents across various lag days. Over the study period, the daily mean PM2.5 mass concentration was 77.8 (standard deviation (SD) = 72.7) µg/m3. We estimated an odds ratio of 1.038 (95% confidence interval (CI): 1.003-1.074), 1.038 (1.013-1.063) and 1.057 (1.023-1.097) for MI mortality associated with per interquartile range (IQR) increase in the 3-day moving-average exposure to PM2.5 (IQR = 76.3 µg/m3), EC (4.1 µg/m3) and OC (9.1 µg/m3), respectively. We did not identify significant association between MI death and exposure to water-soluble ions (SO42-, NH4+ and NO3-). Likelihood ratio tests supported no evident violations of linear assumptions for constituents-MI associations. Subgroup analyses showed stronger associations between MI death and EC/OC exposure in the elderly, males and cold months. Short-term exposure to PM2.5 constituents, particularly those carbonaceous aerosols, was associated with increased risks of MI mortality.

Decoding algae under abnormal temperature variability with odorant as a responsive indicator.

Harmful algal blooms impair the aesthetic quality and healthy performance of in-situ water. Worse yet, dramatic temperature variability arises additional difficulty in algal-induced risk assessment, which is so far poorly explored. Microcystis aeruginosa (FACHB 905), was selected to explore the odor-producing pattern (ß-cyclocitral, the major odorant of the test alga) under several temperature-varying scenarios. Significant differences were observed in total ß-cyclocitral yield between these scenarios, e.g., a rapid yield response as a result of acute temperature variation. Yield response was not only dependent on absolute temperature, but influenced by temperature variability stress. Acute increase (AI) or sequential increase (SI) in temperature caused extra production response, while the opposite was observed in groups with acute decrease (AD) and sequential decrease (SD) in temperature. Cell growth in AD group showed severe inhibition, with the specific growth rates fluctuating around half of that in 16-control. Whereas, SD could relieve such detrimental growth effects. Cell quota of ß-cyclocitral yield was sensitive to temperature variation, with notable increase in AI and SI. Further, peaks in cell quota for SI group (79.3 %) were higher than for AI group (57.9 %). Cell quota variations in temperature-varying conditions contributed to the total yield response (R2 = 0.566-0.980) more than cell intensity variations (R2 = 0.0397-0.548). Further, it was also found that the internal mechanism by reactive oxygen species and pigments varied in various thermal scenarios. Overall, it was demonstrated that more than absolute value differences, temperature varying patterns across time influence algal behavior and related hazards, which should be noted in resource water quality management.

Association of Midday Napping with All-Cause Mortality in Chinese Adults: A 8-Year Nationwide Cohort Study.

This study aimed to investigate the longitudinal association of estimated daytime nap duration with all-cause mortality in Chinese adults. We conceived a prospective cohort design using adult survey data of the baseline and four follow-up waves (2010-2019) from China Family Panel Studies. Cox frailty models with random intercepts for surveyed provinces were used to estimate risks of all-cause mortality associated with midday napping. Trend and subgroup analyses were also performed stratified by demographic, regional and behavioral factors. Compared with non-nappers, those who reported a long napping duration (≥60 min/day) had an increased risk of all-cause mortality, while shorter napping (<60 min) showed no association with mortality. We observed significant trends for greater risks of mortality associated with longer nap duration. Long nap-associated higher risk of all-cause mortality was seen in a group of nocturnal sleep duration ≥9 h. We identified stronger associations of long nap with mortality among adults aged over 50 years, those with lower BMI (<24 kg/m2), residents in rural regions and unregular exercisers. Long midday napping is independently associated with higher risks of all-cause mortality in Chinese adults.

Long-term exposure to PM2.5 constituents in relation to glucose levels and diabetes in middle-aged and older Chinese.

BACKGROUND: Previous studies have indicated the associations between fine particulate matter (PM2.5) exposure and diabetes or glucose levels. However, evidence linking PM2.5 constituents and diabetes or glucose levels was extensively scarce, particularly in developing countries. This study aimed to investigate the associations of exposure to PM2.5 and its five constituents (black carbon [BC], organic matter [OM], nitrate [NO3-], sulfate [SO42-], and ammonium [NH4+]) with diabetes and glucose levels among the middle-aged and elderly Chinese populations. METHODS: A national cross-sectional sample of participants aged 45+ years was enrolled from 28 provinces across China's mainland. Health examination and questionnaire survey for each respondent were performed during 2011-2012. Diabetes was determined by alternative definitions, and the main definition (MD) was self-report diabetes or antidiabetic medicine use or HbA1c ≥6.5 or fasting glucose ≥7 mmol/L or random glucose ≥11.1 mmol/L. Monthly exposure to PM2.5 mass and its five constituents (BC, OM, NO3-, SO42-, and NH4+) for each participant at residence were estimated using satellite-based spatiotemporal prediction models. Generalized linear models and linear mixed-effects models were used to assess the effects of exposure to PM2.5 and its constituents on diabetes or glucose levels, respectively. Stratification analyses were done by sex and age. RESULTS: We included a total of 17,326 adults over 45 years in this study. The 3-year mean (interquartile range [IQR]) concentrations of PM2.5, BC, OM, NO3-, SO42-, and NH4+ were 47.9 (27.4) µg/m3, 2.9 (2.2) µg/m3, 9.2 (6.6) µg/m3, 10.2 (9.4) µg/m3, 11.0 (5.2) µg/m3, and 7.1 (4.4) µg/m3, respectively. Per IQR rise in exposure to PM2.5 was significantly associated with an increase of 0.133 mmol/L (95% confidence interval, 0.048-0.219) in glucose concentrations. Similar positive associations were observed for BC (0.097 mmol/L [0.012-0.181]), OM (0.160 mmol/L [0.065-0.256]), NO3- (0.145 mmol/L [0.039-0.251]), SO42- (0.111 mmol/L [0.026-0.196]), and NH4+ (0.135 mmol/L [0.041-0.230]). Under different diabetes definitions, PM2.5 mass and selected constituents with the exception of SO42- were all associated with a higher risk of prevalent diabetes. In MD-based analysis, similar positive associations were observed for four constituents, with corresponding odds ratios of 1.180 (1.097-1.270) for PM2.5, 1.154 (1.079-1.235) for BC, 1.170 (1.079-1.270) for OM, 1.200 (1.098-1.312) for NO3-, and 1.123 (1.037-1.215) for NH4+. Stratified analyses showed a significantly higher risk of diabetes in males (1.225 [1.064-1.411]) than females (1.024 [0.923-1.136]) when exposed to PM2.5. Participants under 65 years were generally more vulnerable to diabetes hazards related to PM2.5 constituents exposure. CONCLUSIONS: Exposures to PM2.5 and its constituents (i.e., BC, OM, NO3-, and NH4+) were positively associated with increased risks of prevalent diabetes and elevated glucose levels in middle-aged and older adults.

A supervised protein complex prediction method with network representation learning and gene ontology knowledge.

BACKGROUND: Protein complexes are essential for biologists to understand cell organization and function effectively. In recent years, predicting complexes from protein-protein interaction (PPI) networks through computational methods is one of the current research hotspots. Many methods for protein complex prediction have been proposed. However, how to use the information of known protein complexes is still a fundamental problem that needs to be solved urgently in predicting protein complexes. RESULTS: To solve these problems, we propose a supervised learning method based on network representation learning and gene ontology knowledge, which can fully use the information of known protein complexes to predict new protein complexes. This method first constructs a weighted PPI network based on gene ontology knowledge and topology information, reducing the network's noise problem. On this basis, the topological information of known protein complexes is extracted as features, and the supervised learning model SVCC is obtained according to the feature training. At the same time, the SVCC model is used to predict candidate protein complexes from the protein interaction network. Then, we use the network representation learning method to obtain the vector representation of the protein complex and train the random forest model. Finally, we use the random forest model to classify the candidate protein complexes to obtain the final predicted protein complexes. We evaluate the performance of the proposed method on two publicly PPI data sets. CONCLUSIONS: Experimental results show that our method can effectively improve the performance of protein complex recognition compared with existing methods. In addition, we also analyze the biological significance of protein complexes predicted by our method and other methods. The results show that the protein complexes predicted by our method have high biological significance.

Short-term exposure to fine particulate matter constituents and mortality: case-crossover evidence from 32 counties in China.

A growing number of studies associated increased mortality with exposures to specific fine particulate (PM2.5) constituents, while great heterogeneity exists between locations. In China, evidence linking PM2.5 constituents and mortality was extensively sparse. This study primarily aimed to quantify short-term associations between PM2.5 constituents and non-accidental mortality among the Chinese population. We collected daily mortality records from 32 counties in China between January 1, 2011, and December 31, 2013. Daily concentrations of main PM2.5 constituents (organic carbon (OC), elemental carbon (EC), nitrate (NO3-), sulfate (SO42-), and ammonium (NH4+)) were estimated using the modified Community Multiscale Air Quality model. Time-stratified case-crossover design with conditional logistic regression models was adopted to estimate mortality risks associated with short-term exposures to PM2.5 mass and its constituents. Stratification analyses were done by sex, age, and season. A total of 116,959 non-accidental deaths were investigated. PM2.5 concentrations on the day of death were averaged at 75.7 µg m-3 (control day: 75.6 µg m-3), with an interquartile range (IQR) of 65.2 µg m-3. Per IQR rise in PM2.5, EC, OC, NO3-, SO42-, and NH4+ at lag-04 day was associated with an increase in non-accidental mortality of 2.4% (95% confidence interval, (1.0-3.7), 1.7% (0.8-2.7), 2.9% (1.6-4.3), 2.1% (0.4-3.9), 1.0% (0.2-1.9), and 1.6% (0.3-2.9), respectively. Both PM2.5 mass and its constituents were strongly associated with elevated cardiovascular mortality risks, but only PM2.5, EC, and OC were positively associated with respiratory mortality at lag-3 day. PM2.5 mass and its constituents associated effects on mortality varied among sex- and age-specific subpopulations. Differences in the seasonal pattern of associations exist among PM2.5 constituents, with stronger effects related to EC and NO3- in warm months but SO42- and NH4+ in cold months. Short-term exposures to PM2.5 compositions were positively associated with increased risks of mortality, particularly those constituents from combustion-related sources.

Early-life exposure to PM2.5 constituents and childhood asthma and wheezing: Findings from China, Children, Homes, Health study.

BACKGROUND: Emerging evidence suggests that early-life (in-utero and first-year since birth) exposure to ambient PM2.5 is a risk factor for asthma onset and exacerbation among children, while the hazards caused by PM2.5 compositions remain largely unknown. OBJECTIVE: To examine potential associations of early-life exposures to PM2.5 mass and its major chemical constituents with childhood asthma and wheezing. METHODS: By conducting the Phase II of the China, Children, Homes, Health study, we investigated 30,325 preschool children aged 3-6 years during 2019-2020 in mainland China. Early-life exposure to PM2.5 mass and its constituents (i.e., black carbon [BC], organic matter [OM], nitrate, ammonium, sulfate) were calculated based on monthly estimates at a 1 km × 1 km resolution from satellite-based models. We adopted a novel quantile-based g-computation approach to assess the effect of a mixture of PM2.5 constituents on childhood asthma/wheezing. RESULTS: The average PM2.5 concentrations during in-utero and the first year since birth were 64.7 ± 10.6 and 61.8 ± 10.5 µg/m3, respectively. Early-life exposures to a mixture of major PM2.5 constituents were significantly associated with increased risks of asthma and wheezing, while no evident compositions-wheezing associations were found in the first year. Each quintile increases in all five PM2.5 components exposures in utero was accordingly associated with an odds ratio of 1.18 [95% confidence interval: 1.07-1.29] for asthma and 1.08 [1.01-1.16] for wheezing. BC, OM and SO42- contributed more to risks of asthma and wheezing than the other PM2.5 constituents during early life, wherein the effects of BC were only observed during pregnancy. Sex subgroup analyses suggested stronger associations among girls of first-year exposures to PM2.5 components with childhood asthma. CONCLUSION: Early-life exposures to ambient PM2.5, particularly compositions of BC, OM and SO42-, are associated with an increased risk of childhood asthma.

Secular trends in global burden of diabetes attributable to particulate matter pollution from 1990 to 2019.

Increasing evidence suggests an association between fine particulate matter (PM2.5) exposure and type 2 diabetes mellitus. However, there is still a lack of comparative evaluation regarding diabetes burden due to ambient and indoor PM2.5 pollution at a global scale. This study attempts to provide a systematic and comprehensive profile for PM2.5-attributable burden of diabetes and its spatiotemporal trends, globally and regionally. Comparative estimates of diabetes attributable to ambient PM2.5 and household air pollution (HAP) from solid fuels for 204 countries and territories were derived from the Global Burden of Disease Study 2019. Globally, 292.5 (95% uncertainty interval: 207.1, 373.4) thousand deaths and 13.0 (9.1, 17.2) million disability-adjusted life years (DALYs) from diabetes were attributed to PM2.5 pollution in 2019, wherein more than two-thirds (67.3% deaths and 69.7% DALYs) were contributed by ambient PM2.5. Compared to 1990, age-standardized DALY rate (ASDR) in 2019 attributable to ambient PM2.5 increased by 85.9% (APC: 2.21% [2.15, 2.27]), while HAP-associated ASDR decreased by 37.9% (APC: - 1.66% [- 1.82, - 1.50]). We observed a negative correlation between SDI and APC in ASMR (rs = - 0.5, p < 0.001) and ASDR (rs = -0.4, p < 0.001) among 204 countries and territories. HAP-related diabetes experienced a sharp decline during 1990-2019, while global burden of diabetes attributable to ambient PM2.5 was rising rapidly. The elderly and people in low-SDI countries suffered from the greatest burden of diabetes due to PM2.5 pollution. More targeted interventions should be taken by governments to reduce PM2.5 exposure and related diabetes burden.

Longitudinal association of egg intake frequency with cardiovascular disease in Chinese adults.

BACKGROUND AND AIMS: Reducing dietary cholesterol is generally acceptable for the prevention of cardiovascular disease (CVD). Eggs are nutrient-dense and common food items across the world, while rich in cholesterol. The potential effects of egg intake on cardiovascular health remain uncertainty and have been under debate in past decades. METHODS AND RESULTS: A nationwide cohort of 20,688 participants aged 16-110 years without CVD at baseline were derived from the China Family Panel Studies. Egg consumption was assessed by a semi-quantitative food frequency questionnaire. We adopted stratified Cox proportional hazards model with random intercepts for provinces to evaluate associations of egg intake with CVD incidence. During a median follow-up of 6.0 years, we identified 2395 total CVD incidence and mean egg consumption was 3 times/week. Egg intakes were associated lower risks of CVD incidence in the multivariate-adjusted model. Compared with the non-consumers, the corresponding HRs (95% confidence interval) for total CVD events were 0.84 (0.74-0.94) for 1-2 times per week, 0.78 (0.69-0.88) for 3-6/week, and 0.83 (0.72-0.95) for ≥7/week. Similar relationships were found in hypertension. Approximately non-linear relationships were observed between egg consumption with total CVD and hypertension incidence, identifying the lowest risk in 3-6 times/week. Subgroup analyses estimated lower risks of total CVD and hypertension in females only, with significant effect modification by sex (P for interaction = 0.008 and 0.020). CONCLUSION: Egg consumption may be associated with lower risks of CVD incidence among Chinese adults. Our findings could have implications in CVD prevention and might be considered in the development of dietary guidelines.

Longitudinal Impacts of PM2.5 Constituents on Adult Mortality in China.

Limited evidence exists for long-term effects of PM2.5 constituents on mortality. Hence, we aimed to assess associations between all-cause mortality and long-term exposure to PM2.5 constituents in China. We designed a nationwide cohort study of 30524 adults from 162 prefectural areas across mainland China with follow-ups through years 2010-2017. Cox proportional hazards models with time-varying exposures were employed to quantify associations between all-cause mortality and long-term exposure to PM2.5 and constituents. A total of 1210 deaths occurred during 172297.7 person-years. A multiadjusted Cox model estimated an hazard ratio (HR) of 1.125 (95% confidence interval: 1.058-1.197) for all-cause mortality, associated with an interquartile range (IQR = 26.7 µg/m3) rise in exposure to PM2.5. Comparable or stronger associations were found among PM2.5 constituents with the mortality risk increased by 11.3-14.1% per IQR increase in exposure concentrations. After adjustment for the collinearity between total PM2.5 and constituents, effect estimates for nitrate, ammonium, and sulfate remained significant and became larger. Urban residents, alcohol drinkers, smokers, and men were more susceptible to chronic impacts from ambient PM2.5 constituents. This cohort study added the novel longitudinal evidence for elevated mortality linked with long-term exposure to PM2.5 constituents among Chinese adults.