Association between short-term ambient air pollutants and type 2 diabetes outpatient visits: a time series study in Lanzhou, China.

Diabetes is a global public health problem, and the impact of air pollutants on type 2 diabetes mellitus (T2DM) has attracted people's attention. This study aimed to assess the association of short-term exposure to six criteria air pollutants with T2DM outpatient visits in Lanzhou, China. We collected data on daily outpatient visits for T2DM, daily meteorological data and hourly concentrations of air pollutants in Lanzhou from 2013 to 2019. An over-dispersed passion generalized addictive model combined with a distributed lag non-linear model was applied to estimate the associations and stratified analyses were performed by gender, age, and season. The models were fitted with different lag structures, including single lag days from the current to the previous seven days (lag0 to lag7) and moving average concentrations over seven lag days (lag01 to lag07). A positive association between multiple air pollutants, especially PM2.5, NO2, O38h and CO and hospital outpatient visits for T2DM was observed. The largest association between T2DM outpatient visits and PM2.5 was observed at lag06 (RR 1.013, 95% CI: 1.001, 1.027), NO2 at lag03 (RR 1.034, 95% CI: 1.018, 1.050), O38h at lag05 (RR 1.012, 95% CI: 1.001, 1.023) for an increase of 10 µg m-3 and CO at lag03 (RR 1.084, 95% CI: 1.029, 1.142) for an increase of 1 mg m-3 in the concentrations. In addition, people aged <65 and males are more susceptible, and air pollutants have a greater impact on the cold season. This study showed that although the air pollution in Lanzhou was improved, there was still a statistical correlation between air pollution exposure and T2DM outpatient visits. Therefore, the local government still needs to strengthen the control of air pollution and enhance the protection awareness of the diabetic population through education and publicity.

Let the Chart Spark: Embedding Semantic Context into Chart with Text-to-Image Generative Model.

Pictorial visualization seamlessly integrates data and semantic context into visual representation, conveying complex information in an engaging and informative manner. Extensive studies have been devoted to developing authoring tools to simplify the creation of pictorial visualizations. However, mainstream works follow a retrieving-and-editing pipeline that heavily relies on retrieved visual elements from a dedicated corpus, which often compromise data integrity. Text-guided generation methods are emerging, but may have limited applicability due to their predefined entities. In this work, we propose ChartSpark, a novel system that embeds semantic context into chart based on text-to-image generative models. ChartSpark generates pictorial visualizations conditioned on both semantic context conveyed in textual inputs and data information embedded in plain charts. The method is generic for both foreground and background pictorial generation, satisfying the design practices identified from empirical research into existing pictorial visualizations. We further develop an interactive visual interface that integrates a text analyzer, editing module, and evaluation module to enable users to generate, modify, and assess pictorial visualizations. We experimentally demonstrate the usability of our tool, and conclude with a discussion of the potential of using text-to-image generative models combined with an interactive interface for visualization design.

A Natural Compound Containing a Disaccharide Structure of Glucose and Rhamnose Identified as Potential N-Glycanase 1 (NGLY1) Inhibitors.

N-glycanase 1 (NGLY1) is an essential enzyme involved in the deglycosylation of misfolded glycoproteins through the endoplasmic reticulum (ER)-associated degradation (ERAD) pathway, which could hydrolyze N-glycan from N-glycoprotein or N-glycopeptide in the cytosol. Recent studies indicated that NGLY1 inhibition is a potential novel drug target for antiviral therapy. In this study, structure-based virtual analysis was applied to screen candidate NGLY1 inhibitors from 2960 natural compounds. Three natural compounds, Poliumoside, Soyasaponin Bb, and Saikosaponin B2 showed significantly inhibitory activity of NGLY1, isolated from traditional heat-clearing and detoxifying Chinese herbs. Furthermore, the core structural motif of the three NGLY1 inhibitors was a disaccharide structure with glucose and rhamnose, which might exert its action by binding to important active sites of NGLY1, such as Lys238 and Trp244. In traditional Chinese medicine, many compounds containing this disaccharide structure probably targeted NGLY1. This study unveiled the leading compound of NGLY1 inhibitors with its core structure, which could guide future drug development.

SIGANEO: Similarity network with GAN enhancement for immunogenic neoepitope prediction.

Target selection of the personalized cancer neoantigen vaccine, which is highly dependent on computational prediction algorithms, is crucial for its clinical efficacy. Due to the limited number of experimentally validated immunogenic neoepitopes as well as the complexity of neoantigens in eliciting T cell response, the accuracy of neoepitope immunogenicity prediction methods requires persistent efforts for improvement. We present a deep learning framework for neoepitope immunogenicity prediction - SIGANEO by integrating GAN-like network with similarity network to address issues of missing values and limited data concerning neoantigen prediction. This framework exhibits superior performance over competing machine-learning-based neoantigen prediction algorithms over an independent test dataset from TESLA consortium. Particularly for the clinical setting of neoantigen vaccine where only the top 10 and 20 predictions are selected for vaccine production, SIGANEO achieves significantly better accuracy for predicting experimentally validated neoepitopes. Our work demonstrates that deep learning techniques can greatly boost the accuracy of target identification for cancer neoantigen vaccine.

TimeTuner: Diagnosing Time Representations for Time-Series Forecasting with Counterfactual Explanations.

Deep learning (DL) approaches are being increasingly used for time-series forecasting, with many efforts devoted to designing complex DL models. Recent studies have shown that the DL success is often attributed to effective data representations, fostering the fields of feature engineering and representation learning. However, automated approaches for feature learning are typically limited with respect to incorporating prior knowledge, identifying interactions among variables, and choosing evaluation metrics to ensure that the models are reliable. To improve on these limitations, this paper contributes a novel visual analytics framework, namely TimeTuner, designed to help analysts understand how model behaviors are associated with localized correlations, stationarity, and granularity of time-series representations. The system mainly consists of the following two-stage technique: We first leverage counterfactual explanations to connect the relationships among time-series representations, multivariate features and model predictions. Next, we design multiple coordinated views including a partition-based correlation matrix and juxtaposed bivariate stripes, and provide a set of interactions that allow users to step into the transformation selection process, navigate through the feature space, and reason the model performance. We instantiate TimeTuner with two transformation methods of smoothing and sampling, and demonstrate its applicability on real-world time-series forecasting of univariate sunspots and multivariate air pollutants. Feedback from domain experts indicates that our system can help characterize time-series representations and guide the feature engineering processes.

Graph Convolutional Autoencoder and Generative Adversarial Network-Based Method for Predicting Drug-Target Interactions.

The computational prediction of novel drug-target interactions (DTIs) may effectively speed up the process of drug repositioning and reduce its costs. Most previous methods integrated multiple kinds of connections about drugs and targets by constructing shallow prediction models. These methods failed to deeply learn the low-dimension feature vectors for drugs and targets and ignored the distribution of these feature vectors. We proposed a graph convolutional autoencoder and generative adversarial network (GAN)-based method, GANDTI, to predict DTIs. We constructed a drug-target heterogeneous network to integrate various connections related to drugs and targets, i.e., the similarities and interactions between drugs or between targets and the interactions between drugs and targets. A graph convolutional autoencoder was established to learn the network embeddings of the drug and target nodes in a low-dimensional feature space, and the autoencoder deeply integrated different kinds of connections within the network. A GAN was introduced to regularize the feature vectors of nodes into a Gaussian distribution. Severe class imbalance exists between known and unknown DTIs. Thus, we constructed a classifier based on an ensemble learning model, LightGBM, to estimate the interaction propensities of drugs and targets. This classifier completely exploited all unknown DTIs and counteracted the negative effect of class imbalance. The experimental results indicated that GANDTI outperforms several state-of-the-art methods for DTI prediction. Additionally, case studies of five drugs demonstrated the ability of GANDTI to discover the potential targets for drugs.

VISAtlas: An Image-based Exploration and Query System for Large Visualization Collections via Neural Image Embedding.

High-quality visualization collections are beneficial for a variety of applications including visualization reference and data-driven visualization design. The visualization community has created many visualization collections, and developed interactive exploration systems for the collections. However, the systems are mainly based on extrinsic attributes like authors and publication years, whilst neglect intrinsic property (i.e., visual appearance) of visualizations, hindering visual comparison and query of visualization designs. This paper presents VISAtlas, an image-based approach empowered by neural image embedding, to facilitate exploration and query for visualization collections. To improve embedding accuracy, we create a comprehensive collection of synthetic and real-world visualizations, and use it to train a convolutional neural network (CNN) model with a triplet loss for taxonomical classification of visualizations. Next, we design a coordinated multiple view (CMV) system that enables multi-perspective exploration and design retrieval based on visualization embeddings. Specifically, we design a novel embedding overview that leverages contextual layout framework to preserve the context of the embedding vectors with the associated visualization taxonomies, and density plot and sampling techniques to address the overdrawing problem. We demonstrate in three case studies and one user study the effectiveness of VISAtlas in supporting comparative analysis of visualization collections, exploration of composite visualizations, and image-based retrieval of visualization designs. The studies reveal that real-world visualization collections (e.g., Beagle and VIS30K) better accord with the richness and diversity of visualization designs than synthetic collections (e.g., Data2Vis), inspiring composite visualizations are identified in real-world collections, and distinct design patterns exist in visualizations from different sources.

MATHLA: a robust framework for HLA-peptide binding prediction integrating bidirectional LSTM and multiple head attention mechanism.

BACKGROUND: Accurate prediction of binding between class I human leukocyte antigen (HLA) and neoepitope is critical for target identification within personalized T-cell based immunotherapy. Many recent prediction tools developed upon the deep learning algorithms and mass spectrometry data have indeed showed improvement on the average predicting power for class I HLA-peptide interaction. However, their prediction performances show great variability over individual HLA alleles and peptides with different lengths, which is particularly the case for HLA-C alleles due to the limited amount of experimental data. To meet the increasing demand for attaining the most accurate HLA-peptide binding prediction for individual patient in the real-world clinical studies, more advanced deep learning framework with higher prediction accuracy for HLA-C alleles and longer peptides is highly desirable. RESULTS: We present a pan-allele HLA-peptide binding prediction framework-MATHLA which integrates bi-directional long short-term memory network and multiple head attention mechanism. This model achieves better prediction accuracy in both fivefold cross-validation test and independent test dataset. In addition, this model is superior over existing tools regarding to the prediction accuracy for longer ligand ranging from 11 to 15 amino acids. Moreover, our model also shows a significant improvement for HLA-C-peptide-binding prediction. By investigating multiple-head attention weight scores, we depicted possible interaction patterns between three HLA I supergroups and their cognate peptides. CONCLUSION: Our method demonstrates the necessity of further development of deep learning algorithm in improving and interpreting HLA-peptide binding prediction in parallel to increasing the amount of high-quality HLA ligandome data.

Convolutional Neural Network and Bidirectional Long Short-Term Memory-Based Method for Predicting Drug-Disease Associations.

Identifying novel indications for approved drugs can accelerate drug development and reduce research costs. Most previous studies used shallow models for prioritizing the potential drug-related diseases and failed to deeply integrate the paths between drugs and diseases which may contain additional association information. A deep-learning-based method for predicting drug-disease associations by integrating useful information is needed. We proposed a novel method based on a convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM)-CBPred-for predicting drug-related diseases. Our method deeply integrates similarities and associations between drugs and diseases, and paths among drug-disease pairs. The CNN-based framework focuses on learning the original representation of a drug-disease pair from their similarities and associations. As the drug-disease association possibility also depends on the multiple paths between them, the BiLSTM-based framework mainly learns the path representation of the drug-disease pair. In addition, considering that different paths have discriminate contributions to the association prediction, an attention mechanism at path level is constructed. Our method, CBPred, showed better performance and retrieved more real associations in the front of the results, which is more important for biologists. Case studies further confirmed that CBPred can discover potential drug-disease associations.

Inferring Drug-Related Diseases Based on Convolutional Neural Network and Gated Recurrent Unit.

Predicting novel uses for drugs using their chemical, pharmacological, and indication information contributes to minimizing costs and development periods. Most previous prediction methods focused on integrating the similarity and association information of drugs and diseases. However, they tended to construct shallow prediction models to predict drug-associated diseases, which make deeply integrating the information difficult. Further, path information between drugs and diseases is important auxiliary information for association prediction, while it is not deeply integrated. We present a deep learning-based method, CGARDP, for predicting drug-related candidate disease indications. CGARDP establishes a feature matrix by exploiting a variety of biological premises related to drugs and diseases. A novel model based on convolutional neural network (CNN) and gated recurrent unit (GRU) is constructed to learn the local and path representations for a drug-disease pair. The CNN-based framework on the left of the model learns the local representation of the drug-disease pair from their feature matrix. As the different paths have discriminative contributions to the drug-disease association prediction, we construct an attention mechanism at the path level to learn the informative paths. In the right part, a GRU-based framework learns the path representation based on path information between the drug and the disease. Cross-validation results indicate that CGARDP performs better than several state-of-the-art methods. Further, CGARDP retrieves more real drug-disease associations in the top part of the prediction result that are of concern to biologists. Case studies on five drugs demonstrate that CGARDP can discover potential drug-related disease indications.

Gradient Boosting Decision Tree-Based Method for Predicting Interactions Between Target Genes and Drugs.

Determining the target genes that interact with drugs-drug-target interactions-plays an important role in drug discovery. Identification of drug-target interactions through biological experiments is time consuming, laborious, and costly. Therefore, using computational approaches to predict candidate targets is a good way to reduce the cost of wet-lab experiments. However, the known interactions (positive samples) and the unknown interactions (negative samples) display a serious class imbalance, which has an adverse effect on the accuracy of the prediction results. To mitigate the impact of class imbalance and completely exploit the negative samples, we proposed a new method, named DTIGBDT, based on gradient boosting decision trees, for predicting candidate drug-target interactions. We constructed a drug-target heterogeneous network that contains the drug similarities based on the chemical structures of drugs, the target similarities based on target sequences, and the known drug-target interactions. The topological information of the network was captured by random walks to update the similarities between drugs or targets. The paths between drugs and targets could be divided into multiple categories, and the features of each category of paths were extracted. We constructed a prediction model based on gradient boosting decision trees. The model establishes multiple decision trees with the extracted features and obtains the interaction scores between drugs and targets. DTIGBDT is a method of ensemble learning, and it effectively reduces the impact of class imbalance. The experimental results indicate that DTIGBDT outperforms several state-of-the-art methods for drug-target interaction prediction. In addition, case studies on Quetiapine, Clozapine, Olanzapine, Aripiprazole, and Ziprasidone demonstrate the ability of DTIGBDT to discover potential drug-target interactions.

Proteomic characterization of endogenous substrates of mammalian ubiquitin ligase Hrd1.

BACKGROUND: Endoplasmic reticulum (ER)-associated degradation (ERAD) regulates protein homeostasis in the secretory pathway by targeting misfolded or unassembled proteins for degradation by the proteasome. Hrd1 is a conserved multi-spanning membrane bound ubiquitin ligase required for ubiquitination of many aberrant ER proteins, but few endogenous substrates of Hrd1 have been identified to date. METHODS: Using a SILAC-based quantitative proteomic approach combined with CRISPR-mediated gene silencing, we searched for endogenous physiological substrates of Hrd1. We used RNA microarray, immunoblotting, cycloheximide chase combined with chemical genetics to define the role of Hrd1 in regulating the stability of endogenous ERAD substrates. RESULTS: We identified 58 proteins whose levels are consistently upregulated in Hrd1 null HEK293 cells. Many of these proteins function in pathways involved in stress adaptation or immune surveillance. We validated OS9, a lectin required for ERAD of glycoproteins as a highly upregulated protein in Hrd1 deficient cells. Moreover, the abundance of OS9 is inversely correlated with Hrd1 level in clinical synovium samples isolated from osteoarthritis and rheumatoid arthritis patients. Intriguingly, immunoblotting detects two OS9 variants, both of which are upregulated when Hrd1 is inactivated. However, only one of these variants is subject to proteasome dependent degradation that requires Hrd1 and the AAA (ATPase associated with diverse cellular activities) ATPase p97. The stability of the other variant on the other hand is influenced by a lysosomal inhibitor. CONCLUSION: Hrd1 regulates the stability of proteins involved in ER stress response and immune activation by both proteasome dependent and independent mechanisms.

Presynaptic inhibition by α2 receptor/adenylate cyclase/PDE4 complex at retinal rod bipolar synapse.

G-protein-coupled receptor (GPCR)-mediated presynaptic inhibition is a fundamental mechanism regulating synaptic transmission in the CNS. The classical GPCR-mediated presynaptic inhibition in the CNS is produced by direct interactions between the G(ßγ) subunits of the G-protein and presynaptic Ca(2+) channels, K(+) channels, or synaptic proteins that affect transmitter release. This mode of action is shared by well known GPCRs such as the α2, GABA(B), and CB1 receptors. We report that the α2 receptor-mediated inhibition of presynaptic Ca(2+) channel and transmitter release in rat retinal rod bipolar cells depends on the G(α) subunit via a G(α)-adenylate cyclase-cAMP cascade and requires participation of the type 4 phosphodiesterase (PDE4), a new role for phosphodiesterase in neural signaling. By using the G(α) instead of the G(ßγ) subunits, this mechanism is able to use a cyclase/PDE enzyme pair to dynamically control a cyclic nucleotide second messenger (i.e., cAMP) for the regulation of synaptic transmission, an operating strategy that shows remarkable similarity to that of dynamic control of cGMP and transmitter release from photoreceptors by the guanylate cyclase/PDE6 pair in phototransduction. Our results demonstrate a new paradigm of GPCR-mediated presynaptic inhibition in the CNS and add a new regulatory mechanism at a critical presynaptic site in the visual pathway that controls the transmission of scotopic information. They also provide a presynaptic mechanism that could contribute to neuroprotection of retinal ganglion cells by α2 agonists, such as brimonidine, in animal models of glaucoma and retinal ischemia and in glaucoma patients.

[Imaging reasons for periprosthetic femoral fractures during the operation of total hip arthroplasty with anatomic prosthesis].

OBJECTIVE: To explore the imaging reasons for periprosthetic femoral fractures during the operation of total hip arthroplasty with anatomic prosthesis. METHODS: The fracture group consisted of 7 cases with periprosthetic femoral fracture (PPFF) and the non-fracture group 21 cases without PPFF during the operation of total hip arthroplasty (THA) among the 144 cases of primary THA with anatomic prosthesis. The preoperative plain films of hip joint were taken to calibrate the Sharp's angle, centre edge (CE) angle, femoral neck shaft angle, femoral anteversion angle of neck, bowing angle of proximal femoral shaft part, width of narrowest part in proximal femoral shaft. For each case, surgical details, age, height and weight were recorded. The results were analyzed with independent sample t test by software SPSS 17.0. RESULTS: No significant difference existed in general situation, Sharp's angle and CE's angle between two groups. And there were significant differences in femoral neck shaft angle (P = 0.040), femoral anteversion angle of neck (P = 0.026), bowing angle of proximal femoral shaft part (P = 0.024), width of narrowest part in proximal femoral shaft (P = 0.021). CONCLUSION: Dysplasia of proximal femur is a major cause of periprosthetic femoral fracture during the operation of total hip arthroplasty with anatomic prosthesis.

[Dermal hypoesthesia after total knee arthroplasty].

OBJECTIVE: To observe the incidence of skin sensory loss after total knee arthroplasty (TKA) and its natural history over time, and to identify the relationship between numbness area and incision length, tourniquet time, age and gender. METHODS: In the study, 132 patients (20 males and 112 females, with an average age of 69.75 years old, 135 cases of TKA) who underwent primary TKA with midline incisions were chosen and grouped chronologically (4 years, 3 years, 2 years, 1 year, 6 months, 1 month) to the investigation time point from Peking University First Hospital. All the operations were done by the same surgeon team with Stryker NRG and Depuy RP (without patellar resurfacing). Numbness incidence, numbness area, scar length, tourniquet time were recorded from the questionnaires sent to the patients and their medical records. RESULTS: 84.44% of the patients received a reduced skin sensory after TKA, 91.22% of which had a smaller numbness area gradually over time. The numbness area was decreased from the 1 month postoperation group to the 4 years postoperation group (P <0.001). The numbness area in 2 years postoperation group and more were significantly smaller than 1 month postoperation group (P=0.042, 0.004, 0.022), however, the skin flap numbness area had little change after 2 years (P>0.05). The hypoesthesia flap was completely lateral to the incision in 88.60% of the patients, and the numbness area covered the lateral skin and part of media skin to the incision in 11.40% of the patients. Numbness size had no relationship with the patients' gender, age, length of scar and tourniquet time (P>0.05). CONCLUSION: Most but not all the patients have a dermal hypoesthesia after total knee arthroplasty. The numbness area will gradually reduce over time. Numbness size is obviously smaller 2 years postoperation and then it will be stable. Gender, age, length of incision, and tourniquet time have no significant relationship with the size of numbness.

[Treatment of unstable chronic anterior cruciate ligament-deficient knee with osteoarthritis].

OBJECTIVE: Untreated ruptures of anterior cruciate ligament (ACL) lead to progressive meniscus tear and development of knee osteoarthritis over decades. The present study was designed to explore the early results of ACL reconstruction for the patients with unstable ACL-deficient knee with osteoarthritis. METHODS: Twelve patients with a mean age of 46 years (range: 35 - 54) underwent ACL reconstruction for ACL-deficient knee with osteoarthritis. All had chronic anterior instability and one or more episodes of knee instability. There was no previous diagnosis of ACL ruptures and no prior ligament surgery on involved knee. The preoperative duration of symptoms was 1 to 5 years. ACL reconstruction with arthroscopic single-bundle four-strand hamstring tendon autograft was performed for all patients. The laxity of knee was determined with Lachman and pivot tests. The patient subjective evaluation of function was examined with the modified Lysholm scoring scale pre- and post-operatively. RESULTS: Obvious articular cartilage degeneration and cartilage space stenosis in medial compartment were found in 9 patients and complex medial meniscus tear in 10. During the follow-up period of 9 - 36 months, there was no graft failure or loss of ROM (range of motion) for any of these knees. The symptoms of instability were relieved in all patients. The post-operative knee stability improved markedly. The modified Lysholm scores improved from a median pre-operative score of 62.0 points to an average of 89.5 at the last follow-up. CONCLUSION: ACL reconstruction with hamstring tendon may significantly relieve symptoms and improve knee functions in the patients of chronic anterior cruciate ligament-deficient knee with osteoarthritis, especially in those with primary symptoms of instability.