Comparative epigenomic annotation of regulatory DNA.

Despite the explosive growth of genomic data, functional annotation of regulatory sequences remains difficult. Here, we introduce "comparative epigenomics"-interspecies comparison of DNA and histone modifications-as an approach for annotation of the regulatory genome. We measured in human, mouse, and pig pluripotent stem cells the genomic distributions of cytosine methylation, H2A.Z, H3K4me1/2/3, H3K9me3, H3K27me3, H3K27ac, H3K36me3, transcribed RNAs, and P300, TAF1, OCT4, and NANOG binding. We observed that epigenomic conservation was strong in both rapidly evolving and slowly evolving DNA sequences, but not in neutrally evolving sequences. In contrast, evolutionary changes of the epigenome and the transcriptome exhibited a linear correlation. We suggest that the conserved colocalization of different epigenomic marks can be used to discover regulatory sequences. Indeed, seven pairs of epigenomic marks identified exhibited regulatory functions during differentiation of embryonic stem cells into mesendoderm cells. Thus, comparative epigenomics reveals regulatory features of the genome that cannot be discerned from sequence comparisons alone.

Associations among professional quality of life dimensions, burnout, nursing practice environment, and turnover intention in newly graduated nurses.

BACKGROUND: Professional quality of life affects nurses' well-being and the quality of care. However, little attention is paid to the relationships among professional quality of life dimensions, burnout, nursing practice environment, and intention to leave this job in newly graduated nurses. AIMS: To assess the levels of professional quality of life dimensions and turnover intention, to examine the predictors for turnover intention, and to explore the mediating roles of professional quality of life dimensions on the associations between these predictors and turnover intention in Chinese newly graduated nurses. METHODS: This was a cross-sectional study with 315 newly graduated nurses selected from five tertiary hospitals and five secondary hospitals in Sichuan province, China. Multiple regression analysis was used to examine the effects of demographic characteristics and work-related factors on intention to leave this job. Structural equation modeling technique was performed to explore the mediating effect of each domain of professional quality of life on the relationships between the predictors and turnover intention. RESULTS: The prevalence of average levels of burnout, secondary traumatic stress, and compassion satisfaction was 43.2%, 57.1%, and 81.3%, respectively. Moreover, 43.8% and 0.6% of the participants reported high and exceptionally high intention to leave this job. Nursing practice environment, social support, and empathy indirectly and significantly affected turnover intention via the mediating roles of burnout and compassion satisfaction, respectively. However, no significant mediating effect of secondary traumatic stress was found between these predictors and turnover intention. LINKING EVIDENCE TO ACTION: Perceptions of greater nursing practice environment, social support, and empathy result in lower turnover intention via reducing burnout and facilitating compassion satisfaction. Strategies such as developing a supportive work and family environment, and cultivating empathic capacity can be effective methods to mitigate against intention to leave this job in newly graduated nurses.

Possibility of using multiscale normalized difference vegetation index data for the assessment of total suspended solids (TSS) concentrations in surface water: A specific case of scale issues in remote sensing.

The degradation of watersheds creates immense pressure on water quality, especially in arid and semiarid regions. Total suspended solids (TSS) provide essential information to water environmental quality assessments. However, the calibration of direct retrieval models requires complicated preparations and further increases uncertainties. Here, we hypothesized that a common remote sensing index (NDVI, normalized difference vegetation index) could be used to estimate TSS concentrations in water due to the effects of canopy cover. To address this hypothesis, we collected 65 water samples from the Ebinur Lake Watershed, northwest China, to investigate the potential relationships between TSS concentrations and Sentinel-2-based NDVI at various scales (100, 200, 300, 400, and 500 m). Subsequently, we established a classical measurement error (CME) model for the estimation of TSS concentrations. The results showed that TSS concentration is negatively related to the NDVI value at all buffer distances. The 300 m scale mean NDVI value showed the most effective explanation of the variations in TSS concentrations (R2 = 0.83, P-value < 0.001), which indicated that the TSS concentration can be assessed by NDVI. The CME model showed that NDVI values played an important role in the assessment of TSS concentrations in surface water. Furthermore, the results of both leave-one-out cross-validation and the accuracy measure suggested that this specific method is satisfactory. Compared with previous statistical and field monitoring results, the proposed method is promising for cost-effective monitoring of TSS concentrations in water, especially in data-poor watersheds. This specific method may provide the basis for the conservation and management of nonpoint source pollution in arid regions.

EpiAlignment: alignment with both DNA sequence and epigenomic data.

Comparative epigenomics, which subjects both epigenome and genome to interspecies comparison, has become a powerful approach to reveal regulatory features of the genome. Thus elucidated regulatory features surpass the information derived from comparison of genomic sequences alone. Here, we present EpiAlignment, a web-based tool to align genomic regions with both DNA sequence and epigenomic data. EpiAlignment takes DNA sequence and epigenomic profiles derived by ChIP-seq from two species as input data, and outputs the best semi-global alignments. These alignments are based on EpiAlignment scores, computed by a dynamic programming algorithm that accounts for both sequence alignment and epigenome similarity. For timely response, the EpiAlignment web server automatically initiates up to 140 computing threads depending on the size of user input data. For users' convenience, we have pre-compiled the comparable human and mouse epigenome datasets in matched cell types and tissues from the Roadmap Epigenomics and ENCODE consortia. Users can either upload their own data or select pre-compiled datasets as inputs for EpiAlignment analyses. Results are presented in graphical and tabular formats where the entries can be interactively expanded to visualize additional features of these aligned regions. EpiAlignment is available at https://epialign.ucsd.edu/.

The relationships of both transition shock, empathy, resilience and coping strategies with professional quality of life in newly graduated nurses.

BACKGROUND: Data on professional quality of life in newly graduated nurses are scarce. This study aimed to describe the levels of professional quality of life, and to explore the relationships of transition shock, empathy, resilience and coping strategies with professional quality of life in newly graduated nurses. METHODS: This was a cross-sectional study, which used a two-stage sampling method to recruit 393 newly graduated nurses in Sichuan province of China. Multiple regression analysis was used to explore the effects of transition shock, empathy, resilience and coping strategies on professional quality of life. Data were collected using standardized scales. RESULTS: The prevalence of average levels of compassion satisfaction, burnout and secondary traumatic stress in newly graduated nurses were 80.2, 38.2 and 57.5%, respectively. Transition shock was a significant negative predictor, and empathy, resilience and adaptive coping were significant positive predictors for compassion satisfaction. Transition shock and passive coping were significant positive predictors, and empathy was a significant negative predictor for burnout and secondary traumatic stress. Resilience and adaptive coping contributed to burnout significantly and negatively. CONCLUSION: Higher transition shock and lower empathy cause lower compassion satisfaction and higher compassion fatigue. More resilience and adaptive coping cause more compassion satisfaction and less burnout. More passive coping contributes to higher compassion fatigue. Strategies such as transition or preceptorship programmes, and empathy, resilience and coping training are effective methods to reduce transition shock, facilitate empathy, resilience and coping, and consequently, enhance professional quality of life in newly graduated nurses.

Relationships between resilience, empathy, compassion fatigue, work engagement and turnover intention in haemodialysis nurses: A cross-sectional study.

AIM: To explore the relationships between resilience, empathy, compassion fatigue, work engagement and turnover intention in Chinese haemodialysis nurses. BACKGROUND: Although several studies explored nurses' turnover intention in multiple hospital wards, fewer studies focused on turnover intention and its predictors among nurses in dialysis care. METHODS: We conducted a cross-sectional study and adopted a two-stage sampling method to recruit 528 Chinese haemodialysis nurses. Multiple regression analysis was performed to explore the effects of resilience, empathy, compassion fatigue and work engagement on turnover intention. RESULTS: The prevalence of high and exceptionally high levels of turnover intention was 59.1% and 9.0%. Compassion fatigue had the strongest significant effect on turnover intention (ß = 0.276), followed by work engagement (ß = -0.256) and resilience (ß = 0.193). Haemodialysis nurses in tertiary hospitals reported significant higher levels of turnover intention than those in secondary hospitals (ß = 0.127). CONCLUSIONS: Higher levels of compassion fatigue and lower levels of resilience and work engagement can result in higher turnover intention in haemodialysis nurses. IMPLICATIONS FOR NURSING MANAGEMENT: Strategies such as resilience training programme, mindfulness-based intervention and establishing a positive work environment may be effective methods to improve resilience, reduce compassion fatigue, promote work engagement and decrease turnover intention.

Effects of resilience, social support, and work environment on turnover intention in newly graduated nurses: The mediating role of transition shock.

AIM: The aim of this work was to examine the mediating role of transition shock on the relationships between resilience, social support, work environment, and turnover intention in newly graduated nurses. BACKGROUND: Reports assessing the associations between nurses' resilience, social support, work environment, and turnover intention, both directly and through the mediating effect of transition shock, are scarce. METHODS: This was a cross-sectional study, which recruited 361 Chinese newly graduated nurses with <1 year of work experience. RESULTS: Resilience, social support, and work environment directly and significantly predicted transition shock (direct effect -0.158 to -0.350, p < .01). Resilience, work environment, and transition shock directly and significantly affected turnover intention (direct effect -0.118 to -0.257, p < .05). Transition shock mediated the relationships between resilience, social support, work environment, and turnover intention indirectly and significantly (indirect effect -0.019 to -0.041, p < .05). CONCLUSIONS: The results suggest that nurse managers could enhance resilience and social support and establish a positive work environment to reduce transition shock and turnover intention. IMPLICATIONS FOR NURSING MANAGEMENT: Nurse managers could continuously provide resilience training and develop a supportive workplace climate for newly graduated nurses to facilitate smooth transition into clinical practice, could alleviate their transition shock and turnover intention.

A likelihood approach to testing hypotheses on the co-evolution of epigenome and genome.

Central questions to epigenome evolution include whether interspecies changes of histone modifications are independent of evolutionary changes of DNA, and if there is dependence whether they depend on any specific types of DNA sequence changes. Here, we present a likelihood approach for testing hypotheses on the co-evolution of genome and histone modifications. The gist of this approach is to convert evolutionary biology hypotheses into probabilistic forms, by explicitly expressing the joint probability of multispecies DNA sequences and histone modifications, which we refer to as a class of Joint Evolutionary Model for the Genome and the Epigenome (JEMGE). JEMGE can be summarized as a mixture model of four components representing four evolutionary hypotheses, namely dependence and independence of interspecies epigenomic variations to underlying sequence substitutions and to underlying sequence insertions and deletions (indels). We implemented a maximum likelihood method to fit the models to the data. Based on comparison of likelihoods, we inferred whether interspecies epigenomic variations depended on substitution or indels in local genomic sequences based on DNase hypersensitivity and spermatid H3K4me3 ChIP-seq data from human and rhesus macaque. Approximately 5.5% of homologous regions in the genomes exhibited H3K4me3 modification in either species, among which approximately 67% homologous regions exhibited local-sequence-dependent interspecies H3K4me3 variations. Substitutions accounted for less local-sequence-dependent H3K4me3 variations than indels. Among transposon-mediated indels, ERV1 insertions and L1 insertions were most strongly associated with H3K4me3 gains and losses, respectively. By initiating probabilistic formulation on the co-evolution of genomes and epigenomes, JEMGE helps to bring evolutionary biology principles to comparative epigenomic studies.

GeNemo: a search engine for web-based functional genomic data.

A set of new data types emerged from functional genomic assays, including ChIP-seq, DNase-seq, FAIRE-seq and others. The results are typically stored as genome-wide intensities (WIG/bigWig files) or functional genomic regions (peak/BED files). These data types present new challenges to big data science. Here, we present GeNemo, a web-based search engine for functional genomic data. GeNemo searches user-input data against online functional genomic datasets, including the entire collection of ENCODE and mouse ENCODE datasets. Unlike text-based search engines, GeNemo's searches are based on pattern matching of functional genomic regions. This distinguishes GeNemo from text or DNA sequence searches. The user can input any complete or partial functional genomic dataset, for example, a binding intensity file (bigWig) or a peak file. GeNemo reports any genomic regions, ranging from hundred bases to hundred thousand bases, from any of the online ENCODE datasets that share similar functional (binding, modification, accessibility) patterns. This is enabled by a Markov Chain Monte Carlo-based maximization process, executed on up to 24 parallel computing threads. By clicking on a search result, the user can visually compare her/his data with the found datasets and navigate the identified genomic regions. GeNemo is available at www.genemo.org.

Improved Physicochemical Properties of Yogurt Fortified with Fish Oil/γ-Oryzanol by Nanoemulsion Technology.

Fish oil has several dietary benefits, but its application in food formulations is limited because of its poor water-solubility, easy oxidation and strong odor. The purposes of this study were to produce a fish oil/γ-oryzanol nanoemulsion and to evaluate the effect of adding this nanoemulsion on the physicochemical and sensory characteristics of yogurts. Adding fish oil/γ-oryzanol nanoemulsion resulted in a significant reduction in the acidity and syneresis of yogurt. Yogurt with the nanoemulsion had significantly lower peroxide value (0.28 mmol/L after 21 days) and higher retention of eicosapentaenoic acid and docosahexaenoic acid contents (decreased to 95% and 94% of its initial value, respectively) than yogurt with fish oil/γ-oryzanol (peroxide value = 0.65 mmol/L; eicosapentaenoic acid and docosahexaenoic acid contents decreased to 72% and 53% of its initial value, respectively). Fish oil/γ-oryzanol nanoemulsion incorporated into yogurt had closer sensory attributes scores to plain yogurt. This study may have important implications for the application of fish oil/γ-oryzanol nanoemulsion in yogurt.

Transcriptomic insights into the genetic basis of mammalian limb diversity.

BACKGROUND: From bat wings to whale flippers, limb diversification has been crucial to the evolutionary success of mammals. We performed the first transcriptome-wide study of limb development in multiple species to explore the hypothesis that mammalian limb diversification has proceeded through the differential expression of conserved shared genes, rather than by major changes to limb patterning. Specifically, we investigated the manner in which the expression of shared genes has evolved within and among mammalian species. RESULTS: We assembled and compared transcriptomes of bat, mouse, opossum, and pig fore- and hind limbs at the ridge, bud, and paddle stages of development. Results suggest that gene expression patterns exhibit larger variation among species during later than earlier stages of limb development, while within species results are more mixed. Consistent with the former, results also suggest that genes expressed at later developmental stages tend to have a younger evolutionary age than genes expressed at earlier stages. A suite of key limb-patterning genes was identified as being differentially expressed among the homologous limbs of all species. However, only a small subset of shared genes is differentially expressed in the fore- and hind limbs of all examined species. Similarly, a small subset of shared genes is differentially expressed within the fore- and hind limb of a single species and among the forelimbs of different species. CONCLUSIONS: Taken together, results of this study do not support the existence of a phylotypic period of limb development ending at chondrogenesis, but do support the hypothesis that the hierarchical nature of development translates into increasing variation among species as development progresses.

Cell fate inclination within 2-cell and 4-cell mouse embryos revealed by single-cell RNA sequencing.

It remains an open question when and how the first cell fate decision is made in mammals. Using deep single-cell RNA-seq of matched sister blastomeres, we report highly reproducible inter-blastomere differences among 10 2-cell and five 4-cell mouse embryos. Inter-blastomere gene expression differences dominated between-embryo differences and noise, and were sufficient to cluster sister blastomeres into distinct groups. Dozens of protein-coding genes exhibited reproducible bimodal expression in sister blastomeres, which cannot be explained by random fluctuations. The protein expression of one gene out of four of these bimodal genes tested, Gadd45a, exhibited clear inter-blastomeric contrasts. We traced some of the bimodal mRNA expressions to embryonic genome activation, and others to blastomere-specific RNA depletion. Inter-blastomere differences created coexpression gene networks that were much stronger and larger than those that can possibly be created by random noise. The highly correlated gene pairs at the 4-cell stage overlapped with those showing the same directions of differential expression between inner cell mass (ICM) and trophectoderm (TE). These data substantiate the hypothesis of inter-blastomere differences in 2- and 4-cell mouse embryos, and associate these differences with ICM/TE differences.

Time-variant clustering model for understanding cell fate decisions.

Both spatial characteristics and temporal features are often the subjects of concern in physical, social, and biological studies. This work tackles the clustering problems for time course data in which the cluster number and clustering structure change with respect to time, dubbed time-variant clustering. We developed a hierarchical model that simultaneously clusters the objects at every time point and describes the relationships of the clusters between time points. The hidden layer of this model is a generalized form of branching processes. A reversible-jump Markov Chain Monte Carlo method was implemented for model inference, and a feature selection procedure was developed. We applied this method to explore an open question in preimplantation embryonic development. Our analyses using single-cell gene expression data suggested that the earliest cell fate decision could start at the 4-cell stage in mice, earlier than the commonly thought 8- to 16-cell stage. These results together with independent experimental data from single-cell RNA-seq provided support against a prevailing hypothesis in mammalian development.

Symptom clusters in patients receiving haemodialysis: a systematic review of observational studies.

AIMS AND OBJECTIVES: To systematically review the common composition and the analytical methods performed to derive symptom clusters in patients receiving haemodialysis, and to examine their common predictive and outcome variables. BACKGROUND: Patients receiving haemodialysis may suffer from multiple and interactive symptoms simultaneously, which may impact their mortality, morbidity and quality of life. The composition of these symptom clusters varies depending on the different assessment tools, the variant evaluation timing points and the various statistical methodologies. A detailed review of existing studies is needed. DESIGN: A systematic review of observational studies. METHODS: Medline from 1950, CINAHL from 1960, Embase from 1980 and PsycINFO from 1967 as well as additional sources were searched. Content analysis was conducted to identify articles assessing the interrelationships of multiple symptoms. RESULTS: Five studies were identified. Several symptom clusters were identified with five common groupings being uraemic cluster, neuromuscular cluster, skin cluster, gastrointestinal cluster and energy/fatigue cluster. Four studies performed the principal component analysis with varimax rotation to extract symptom clusters and one study used correlation analysis. Different symptom assessment tools were used and each involving different array of symptoms. The predictive and outcome variables of symptom clusters also varied considerably. No studies have examined the longitudinal course and patients' subjective experiences of symptom clusters. CONCLUSIONS: Inconsistencies in the composition of symptom clusters across studies were identified due to inconsistencies in symptom assessment tools and statistical methodologies. Future studies should focus on an agreement about a robust and clinically relevant definition on symptom clusters; a multidimensional, valid and reliable symptom assessment tool; and an optimal analytical method in patients receiving haemodialysis. RELEVANCE TO CLINICAL PRACTICE: Knowledge of symptom clusters may contribute to understanding the aetiology and pathophysiology of multiple concurrent and interactive symptoms in patients receiving haemodialysis and exploring the effects of these symptoms on clinical outcomes.

Probable Post-Traumatic Stress Disorder and Its Predictors in Disaster-Bereaved Survivors: A Longitudinal Study After the Sichuan Earthquake.

This study examined the trajectory of probable PTSD prevalence and severity, and analyzed the predictors for PTSD severity in bereaved survivors at 6 months and 18 months after the 2008 Sichuan earthquake. This was a longitudinal study with 226 bereaved survivors sampled at 6 months and 18 months post-earthquake. The instrument used in the study was the revised version of the Impact of Event Scale. The results showed that the prevalence of probable PTSD in bereaved survivors decreased significantly from 38.9% at 6 months to 16.8% at 18 months post-earthquake. Loss of a child, being directly exposed to the death of family members and property loss during the earthquake, and mental health services utilization after the earthquake were significant predictors for PTSD severity at both assessments. These findings can contribute to post-disaster psychological rescue work. The bereaved survivors at high risk for more severe PTSD should be particularly targeted.

The effect of perceived organisational support on burnout among community health nurses in China: the mediating role of professional self-concept.

AIM: To examine the mediating effect of professional self-concept on the association between perceived organisational support and burnout among community health nurses in Chengdu, China. BACKGROUND: Burnout is a common phenomenon among nurses and previous studies have focused on work environmental factors contributing to burnout. Limited studies have examined the effects of perceived organisational support and professional self-concept on burnout among community health nurses. METHODS: This was a cross-sectional study with 551 community health nurses in Chengdu, China, which included a two-stage sampling method. Structural equation modelling was used to examine the relationships among perceived organisational support, professional self-concept and burnout. RESULTS: The final sample included 456 nurses (82.7%). Perceived organisational support was a significant positive direct predictor for professional self-concept and a significant negative direct predictor for burnout. Professional self-concept was a significant negative direct contributor to burnout. Professional self-concept had a mediating effect on the relationship between perceived organisational support and burnout. CONCLUSIONS: Perceived organisational support may result in reduced burnout by facilitating the development of positive professional self-concept. IMPLICATIONS FOR NURSING MANAGEMENT: Strategies such as establishing a supportive work environment and professional competence training may be effective methods for burnout prevention and management among community health nurses.

Effect of professional self-concept on burnout among community health nurses in Chengdu, China: the mediator role of organisational commitment.

AIMS AND OBJECTIVES: To examine the associations among professional self-concept, organisational commitment and burnout, and to analyse the mediating role of organisational commitment on the relationship between professional self-concept and burnout among community health nurses in Chengdu, China. BACKGROUND: Previous studies have focused on work environmental variables that contributed to burnout in nurses. However, no study has explored the mediating effect of organisational commitment on the correlation between professional self-concept and burnout in community health nurses. DESIGN: A cross-sectional descriptive study. METHODS: This study was conducted at 36 community health centres in Chengdu, China with 485 nurses sampled using a two-stage sampling method. The measures used in our study included Nurses' Self-concept Questionnaire, Organisational Commitment Scale and Maslach Burnout Inventory. RESULTS: The results of structural equation model techniques indicated that, in the direct approach, positive professional self-concept resulted in increased organisational commitment and reduced burnout. Higher organisational commitment resulted in less burnout. In the indirect approach, organisational commitment performed as a partial mediator on the correlation between professional self-concept and burnout. CONCLUSIONS: Positive perception of professional self-concept can result in reduced burnout via enhancing organisational commitment. RELEVANCE TO CLINICAL PRACTICE: It is crucial for nursing administrators to develop effective intervention strategies such as skills escalator training and assertive training, and establishing a supportive working environment to enhance nurses' professional self-concept and organisational commitment, and decrease burnout.

Enabling interspecies epigenomic comparison with CEpBrowser.

SUMMARY: We developed the Comparative Epigenome Browser (CEpBrowser) to allow the public to perform multi-species epigenomic analysis. The web-based CEpBrowser integrates, manages and visualizes sequencing-based epigenomic datasets. Five key features were developed to maximize the efficiency of interspecies epigenomic comparisons. AVAILABILITY: CEpBrowser is a web application implemented with PHP, MySQL, C and Apache. URL: http://www.cepbrowser.org/.

Understanding variation in transcription factor binding by modeling transcription factor genome-epigenome interactions.

Despite explosive growth in genomic datasets, the methods for studying epigenomic mechanisms of gene regulation remain primitive. Here we present a model-based approach to systematically analyze the epigenomic functions in modulating transcription factor-DNA binding. Based on the first principles of statistical mechanics, this model considers the interactions between epigenomic modifications and a cis-regulatory module, which contains multiple binding sites arranged in any configurations. We compiled a comprehensive epigenomic dataset in mouse embryonic stem (mES) cells, including DNA methylation (MeDIP-seq and MRE-seq), DNA hydroxymethylation (5-hmC-seq), and histone modifications (ChIP-seq). We discovered correlations of transcription factors (TFs) for specific combinations of epigenomic modifications, which we term epigenomic motifs. Epigenomic motifs explained why some TFs appeared to have different DNA binding motifs derived from in vivo (ChIP-seq) and in vitro experiments. Theoretical analyses suggested that the epigenome can modulate transcriptional noise and boost the cooperativity of weak TF binding sites. ChIP-seq data suggested that epigenomic boost of binding affinities in weak TF binding sites can function in mES cells. We showed in theory that the epigenome should suppress the TF binding differences on SNP-containing binding sites in two people. Using personal data, we identified strong associations between H3K4me2/H3K9ac and the degree of personal differences in NFκB binding in SNP-containing binding sites, which may explain why some SNPs introduce much smaller personal variations on TF binding than other SNPs. In summary, this model presents a powerful approach to analyze the functions of epigenomic modifications. This model was implemented into an open source program APEG (Affinity Prediction by Epigenome and Genome, http://systemsbio.ucsd.edu/apeg).

Prediction of PM2.5 concentration based on a CNN-LSTM neural network algorithm.

Fine particulate matter (PM2.5) is a major air pollutant affecting human survival, development and health. By predicting the spatial distribution concentration of PM2.5, pollutant sources can be better traced, allowing measures to protect human health to be implemented. Thus, the purpose of this study is to predict and analyze the PM2.5 concentration of stations based on the integrated deep learning of a convolutional neural network long short-term memory (CNN-LSTM) model. To solve the complexity and nonlinear characteristics of PM2.5 time series data problems, we adopted the CNN-LSTM deep learning model. We collected the PM2.5data of Qingdao in 2020 as well as meteorological factors such as temperature, wind speed and air pressure for pre-processing and characteristic analysis. Then, the CNN-LSTM deep learning model was integrated to capture the temporal and spatial features and trends in the data. The CNN layer was used to extract spatial features, while the LSTM layer was used to learn time dependencies. Through comparative experiments and model evaluation, we found that the CNN-LSTM model can achieve excellent PM2.5 prediction performance. The results show that the coefficient of determination (R2) is 0.91, and the root mean square error (RMSE) is 8.216 µg/m3. The CNN-LSTM model achieves better prediction accuracy and generalizability compared with those of the CNN and LSTM models (R2 values of 0.85 and 0.83, respectively, and RMSE values of 11.356 and 14.367, respectively). Finally, we analyzed and explained the predicted results. We also found that some meteorological factors (such as air temperature, pressure, and wind speed) have significant effects on the PM2.5 concentration at ground stations in Qingdao. In summary, by using deep learning methods, we obtained better prediction performance and revealed the association between PM2.5 concentration and meteorological factors. These findings are of great significance for improving the quality of the atmospheric environment and protecting public health.