RESUMO
Constructing accurate gene regulatory network s (GRNs), which reflect the dynamic governing process between genes, is critical to understanding the diverse cellular process and unveiling the complexities in biological systems. With the development of computer sciences, computational-based approaches have been applied to the GRNs inference task. However, current methodologies face challenges in effectively utilizing existing topological information and prior knowledge of gene regulatory relationships, hindering the comprehensive understanding and accurate reconstruction of GRNs. In response, we propose a novel graph neural network (GNN)-based Multi-Task Learning framework for GRN reconstruction, namely MTLGRN. Specifically, we first encode the gene promoter sequences and the gene biological features and concatenate the corresponding feature representations. Then, we construct a multi-task learning framework including GRN reconstruction, Gene knockout predict, and Gene expression matrix reconstruction. With joint training, MTLGRN can optimize the gene latent representations by integrating gene knockout information, promoter characteristics, and other biological attributes. Extensive experimental results demonstrate superior performance compared with state-of-the-art baselines on the GRN reconstruction task, efficiently leveraging biological knowledge and comprehensively understanding the gene regulatory relationships. MTLGRN also pioneered attempts to simulate gene knockouts on bulk data by incorporating gene knockout information.
Assuntos
Biologia Computacional , Redes Reguladoras de Genes , Biologia Computacional/métodos , Técnicas de Inativação de Genes , Redes Neurais de Computação , Humanos , Regiões Promotoras Genéticas , AlgoritmosRESUMO
Here, we present the manually curated Global Catalogue of Pathogens (gcPathogen), an extensive genomic resource designed to facilitate rapid and accurate pathogen analysis, epidemiological exploration and monitoring of antibiotic resistance features and virulence factors. The catalogue seamlessly integrates and analyzes genomic data and associated metadata for human pathogens isolated from infected patients, animal hosts, food and the environment. The pathogen list is supported by evidence from medical or government pathogenic lists and publications. The current version of gcPathogen boasts an impressive collection of 1 164 974 assemblies comprising 986 044 strains from 497 bacterial taxa, 4794 assemblies encompassing 4319 strains from 265 fungal taxa, 89 965 assemblies featuring 13 687 strains from 222 viral taxa, and 646 assemblies including 387 strains from 159 parasitic taxa. Through this database, researchers gain access to a comprehensive 'one-stop shop' that facilitates global, long-term public health surveillance while enabling in-depth analysis of genomes, sequence types, antibiotic resistance genes, virulence factors and mobile genetic elements across different countries, diseases and hosts. To access and explore the data and statistics, an interactive web interface has been developed, which can be accessed at https://nmdc.cn/gcpathogen/. This user-friendly platform allows seamless querying and exploration of the extensive information housed within the gcPathogen database.
Assuntos
Bases de Dados Genéticas , Infecções , Saúde Pública , Humanos , Genoma Bacteriano/genética , Genômica , Fatores de Virulência/genética , Infecções/microbiologia , Infecções/parasitologia , Infecções/virologia , AnimaisRESUMO
Drug-drug interaction (DDI) prediction can discover potential risks of drug combinations in advance by detecting drug pairs that are likely to interact with each other, sparking an increasing demand for computational methods of DDI prediction. However, existing computational DDI methods mostly rely on the single-view paradigm, failing to handle the complex features and intricate patterns of DDIs due to the limited expressiveness of the single view. To this end, we propose a Hierarchical Triple-view Contrastive Learning framework for Drug-Drug Interaction prediction (HTCL-DDI), leveraging the molecular, structural and semantic views to model the complicated information involved in DDI prediction. To aggregate the intra-molecular compositional and structural information, we present a dual attention-aware network in the molecular view. Based on the molecular view, to further capture inter-molecular information, we utilize the one-hop neighboring information and high-order semantic relations in the structural view and semantic view, respectively. Then, we introduce contrastive learning to enhance drug representation learning from multifaceted aspects and improve the robustness of HTCL-DDI. Finally, we conduct extensive experiments on three real-world datasets. All the experimental results show the significant improvement of HTCL-DDI over the state-of-the-art methods, which also demonstrates that HTCL-DDI opens new avenues for ensuring medication safety and identifying synergistic drug combinations.
Assuntos
Aprendizado Profundo , Interações Medicamentosas , SemânticaRESUMO
Commitment to specific cell lineages is critical for mammalian embryonic development. Lineage determination, differentiation, maintenance, and organogenesis result in diverse life forms composed of multiple cell types. To understand the formation and maintenance of living individuals, including human beings, a comprehensive database that integrates multi-omic information underlying lineage differentiation across multiple species is urgently needed. Here, we construct Lineage Landscape, a database that compiles, analyzes and visualizes transcriptomic and epigenomic information related to lineage development in a collection of species. This landscape draws together datasets that capture the ongoing changes in cell lineages from classic model organisms to human beings throughout embryonic, fetal, adult, and aged stages, providing comprehensive, open-access information that is useful to researchers of a broad spectrum of life science disciplines. Lineage Landscape contains single-cell gene expression and bulk transcriptomic, DNA methylation, histone modifications, and chromatin accessibility profiles. Using this database, users can explore genes of interest that exhibit dynamic expression patterns at the transcriptional or epigenetic levels at different stages of lineage development. Lineage Landscape currently includes over 6.6 million cells, 15 million differentially expressed genes and 36 million data entries across 10 species and 34 organs. Lineage Landscape is free to access, browse, search, and download at http://data.iscr.ac.cn/lineage/#/home.
Assuntos
Linhagem da Célula , Mamíferos , Animais , Humanos , Diferenciação Celular , Cromatina/genética , Bases de Dados Factuais , Metilação de DNA , Mamíferos/genética , Mamíferos/crescimento & desenvolvimento , Expressão GênicaRESUMO
The genomic variations of SARS-CoV-2 continue to emerge and spread worldwide. Some mutant strains show increased transmissibility and virulence, which may cause reduced protection provided by vaccines. Thus, it is necessary to continuously monitor and analyze the genomic variations of SARS-COV-2 genomes. We established an evaluation and prewarning system, SARS-CoV-2 variations evaluation and prewarning system (VarEPS), including known and virtual mutations of SARS-CoV-2 genomes to achieve rapid evaluation of the risks posed by mutant strains. From the perspective of genomics and structural biology, the database comprehensively analyzes the effects of known variations and virtual variations on physicochemical properties, translation efficiency, secondary structure, and binding capacity of ACE2 and neutralizing antibodies. An AI-based algorithm was used to verify the effectiveness of these genomics and structural biology characteristic quantities for risk prediction. This classifier could be further used to group viral strains by their transmissibility and affinity to neutralizing antibodies. This unique resource makes it possible to quickly evaluate the variation risks of key sites, and guide the research and development of vaccines and drugs. The database is freely accessible at www.nmdc.cn/ncovn.
Assuntos
COVID-19/virologia , Bases de Dados Factuais , Mutação , SARS-CoV-2/genética , Algoritmos , Enzima de Conversão de Angiotensina 2/metabolismo , Anticorpos Neutralizantes/metabolismo , Inteligência Artificial , Primers do DNA , Genoma Viral , HumanosRESUMO
Taxonomic and functional research of microorganisms has increasingly relied upon genome-based data and methods. As the depository of the Global Catalogue of Microorganisms (GCM) 10K prokaryotic type strain sequencing project, Global Catalogue of Type Strain (gcType) has published 1049 type strain genomes sequenced by the GCM 10K project which are preserved in global culture collections with a valid published status. Additionally, the information provided through gcType includes >12 000 publicly available type strain genome sequences from GenBank incorporated using quality control criteria and standard data annotation pipelines to form a high-quality reference database. This database integrates type strain sequences with their phenotypic information to facilitate phenotypic and genotypic analyses. Multiple formats of cross-genome searches and interactive interfaces have allowed extensive exploration of the database's resources. In this study, we describe web-based data analysis pipelines for genomic analyses and genome-based taxonomy, which could serve as a one-stop platform for the identification of prokaryotic species. The number of type strain genomes that are published will continue to increase as the GCM 10K project increases its collaboration with culture collections worldwide. Data of this project is shared with the International Nucleotide Sequence Database Collaboration. Access to gcType is free at http://gctype.wdcm.org/.
Assuntos
Bases de Dados Genéticas , Genoma , Filogenia , Células Procarióticas/metabolismo , Pesquisa , Sequência de Bases , Análise de Dados , RNA Ribossômico 16S/genéticaRESUMO
The China National Center for Food Safety Risk Assessment (CFSA) uses the Foodborne Disease Monitoring and Reporting System (FDMRS) to monitor outbreaks of foodborne diseases across the country. However, there are problems of underreporting or erroneous reporting in FDMRS, which significantly increase the cost of related epidemic investigations. To solve this problem, we designed a model to identify suspected outbreaks from the data generated by the FDMRS of CFSA. In this study, machine learning models were used to fit the data. The recall rate and F1-score were used as evaluation metrics to compare the classification performance of each model. Feature importance and pathogenic factors were identified and analyzed using tree-based and gradient boosting models. Three real foodborne disease outbreaks were then used to evaluate the best performing model. Furthermore, the SHapley Additive exPlanation value was used to identify the effect of features. Among all machine learning classification models, the eXtreme Gradient Boosting (XGBoost) model achieved the best performance, with the highest recall rate and F1-score of 0.9699 and 0.9582, respectively. In terms of model validation, the model provides a correct judgment of real outbreaks. In the feature importance analysis with the XGBoost model, the health status of the other people with the same exposure has the highest weight, reaching 0.65. The machine learning model built in this study exhibits high accuracy in recognizing foodborne disease outbreaks, thus reducing the manual burden for medical staff. The model helped us identify the confounding factors of foodborne disease outbreaks. Attention should be paid not only to the health status of those with the same exposure but also to the similarity of the cases in time and space.
Assuntos
Surtos de Doenças/estatística & dados numéricos , Doenças Transmitidas por Alimentos/epidemiologia , Análise de Perigos e Pontos Críticos de Controle/métodos , Aprendizado de Máquina , Vigilância da População/métodos , China/epidemiologia , Doenças Transmitidas por Alimentos/microbiologia , Humanos , Medição de Risco/métodosRESUMO
Central Environmental Inspection (CEI) is a particularly important innovative strategy in the transition of environmental governance in China. The first round of CEI for all provincial regions in mainland China has been finished by the end of 2017, but its actual performance remains to be seen. In this study, a multi-dimensional index system was developed under the framework of Balanced Scorecard. Using the content analysis method, we comprehensively evaluated the performance of CEIs in all provinces inspected from the perspectives of target achievement, local rectification, direct effect, and social involvement. The results indicate that CEI has made encouraging progress in the area of environmental governance and the accumulated experiences of the inspections in the early stage greatly boosted the subsequent performance of CEIs. The provincial performance of the central region was significantly higher than that of other regions. For target achievement, the focal points has been basically realized. Despite some neglect of CEI feedback, local environmental governance is experiencing a promising shift from passive to active in general. For social involvement, the CEI has not only promoted the awakening of public environmental consciousness, but also driven public participation in environmental protection. It is notable that the implementation of environmental co-responsibility between Party and governmental officials needs to be further improved. In addition, the shortcomings of each province were identified as well and policy recommendations for existing problems were offered to guide future optimization of local environmental governance and CEI practice.
Assuntos
Conservação dos Recursos Naturais , Política Ambiental , China , Participação da Comunidade , HumanosRESUMO
With the deployment of multi-modality and large-scale sensor networks for monitoring air quality, we are now able to collect large and multi-dimensional spatio-temporal datasets. For these sensed data, we present a comprehensive visual analysis approach for air quality analysis. This approach integrates several visual methods, such as map-based views, calendar views, and trends views, to assist the analysis. Among those visual methods, map-based visual methods are used to display the locations of interest, and the calendar and the trends views are used to discover the linear and periodical patterns. The system also provides various interaction tools to combine the map-based visualization, trends view, calendar view and multi-dimensional view. In addition, we propose a self-adaptive calendar-based controller that can flexibly adapt the changes of data size and granularity in trends view. Such a visual analytics system would facilitate big-data analysis in real applications, especially for decision making support.
RESUMO
Mercury emission from industrial wastewater has a great impact on the aquatic environment but is not well studied. Inventory analysis, decoupling and decomposition methods have been conducted based on the China Pollution Source Census dataset, which combines industry removal efficiencies to calculate mercury emissions from industrial wastewater in 340 cities in China during 2000-2010. The results show that over these 11 years, total mercury emissions and per capita mercury emissions increased by approximately 5 times, while the emission intensity increased by only about 3%. From 2000 to 2010, only 0.59% of cities showed strong decoupling between economic growth and mercury emissions, and 37.65% of cities showed weak decoupling, whereas 38.82% of cities showed negative decoupling. We attribute the decoupling of economic development and emissions in individual cities to several socioeconomic factors and find that a decline in emission intensity is the main driver. The Gini coefficient indicates a significant imbalance between cities' emissions, but this situation improved during 2000-2010. The objective of this article is to provide a historical perspective on the situation of mercury emissions from wastewater in China, thereby contributing' to the broader understanding of industrial pollution.
Assuntos
Desenvolvimento Econômico , Mercúrio , Humanos , Cidades , Águas Residuárias , Indústrias , China , Dióxido de Carbono/análise , Carbono/análiseRESUMO
Water and energy are two essential resources of modern civilization and are inherently linked. Indeed, the optimization of the water supply system would reduce energy demands and greenhouse gas emissions in the municipal water sector. This research measured the climatic cobenefit of water conservation based on a water flow analysis. The results showed that the estimated energy consumption of the total water system in Changzhou, China, reached approximately 10% of the city's total energy consumption, whereas the industrial sector was found to be more energy intensive than other sectors within the entire water system, accounting for nearly 70% of the total energy use of the water system. In addition, four sustainable water management scenarios would bring the cobenefit of reducing the total energy use of the water system by 13.9%, and 77% of the energy savings through water conservation was indirect. To promote sustainable water management and reduce greenhouse gas emissions, China would require its water price system, both for freshwater and recycled water, to be reformed.
Assuntos
Conservação dos Recursos Naturais , Efeito Estufa/prevenção & controle , Abastecimento de Água , Poluição do Ar/prevenção & controle , China , Cidades , Chuva , Reciclagem , Águas ResiduáriasRESUMO
Promoting the use of reusable takeaway food container (RTFC) in takeaway industry is an effective way to reduce the negative environmental impacts caused by single-use plastic containers. This study intended to figure out the barriers to the new business model deployment through evaluating the economic costs and benefits of RTFC from a stakeholder's perspective. Taking the pilot RTFC project at a university in Guangdong province as a case, we established a holistic cost and benefit analysis framework from a stakeholder's perspective. Both the costs and benefits with and without a market price of each stakeholder were evaluated using market price method and contingent valuation method. The analysis result shows that while shifting to reusable takeaway food container, the costs and benefits of all the main stakeholders changed. The net benefit of consumers is positive about 360 thousand yuan during 2020-2025, while the platform company, the university and the restaurants gain negative net benefits ranging from - 20 to - 470 thousand yuan under current operation situation, which may hinder the sustainable development of this new business model. However, the sensitivity analysis shows that all the stakeholders could gain a positive net benefit by adjusting the rental price, cleaning price and packaging price, as well as optimizing the location of recycling cabinets.
Assuntos
Embalagem de Alimentos , Indústrias , Humanos , Análise Custo-Benefício , China , Desenvolvimento SustentávelRESUMO
Multimodal medical images are widely used by clinicians and physicians to analyze and retrieve complementary information from high-resolution images in a non-invasive manner. Loss of corresponding image resolution adversely affects the overall performance of medical image interpretation. Deep learning-based single image super resolution (SISR) algorithms have revolutionized the overall diagnosis framework by continually improving the architectural components and training strategies associated with convolutional neural networks (CNN) on low-resolution images. However, existing work lacks in two ways: i) the SR output produced exhibits poor texture details, and often produce blurred edges, ii) most of the models have been developed for a single modality, hence, require modification to adapt to a new one. This work addresses (i) by proposing generative adversarial network (GAN) with deep multi-attention modules to learn high-frequency information from low-frequency data. Existing approaches based on the GAN have yielded good SR results; however, the texture details of their SR output have been experimentally confirmed to be deficient for medical images particularly. The integration of wavelet transform (WT) and GANs in our proposed SR model addresses the aforementioned limitation concerning textons. While the WT divides the LR image into multiple frequency bands, the transferred GAN uses multi-attention and upsample blocks to predict high-frequency components. Additionally, we present a learning method for training domain-specific classifiers as perceptual loss functions. Using a combination of multi-attention GAN loss and a perceptual loss function results in an efficient and reliable performance. Applying the same model for medical images from diverse modalities is challenging, our work addresses (ii) by training and performing on several modalities via transfer learning. Using two medical datasets, we validate our proposed SR network against existing state-of-the-art approaches and achieve promising results in terms of structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR).
RESUMO
The electrocatalytic CO2 reduction process has gained enormous attention for both environmental protection and chemicals production. Thereinto, the design of new electrocatalysts with high activity and selectivity can draw inspiration from the abundant scientific literature. An annotated and verified corpus made from massive literature can assist the development of natural language processing (NLP) models, which can offer insight to help guide the understanding of these underlying mechanisms. To facilitate data mining in this direction, we present a benchmark corpus of 6,086 records manually extracted from 835 electrocatalytic publications, along with an extended corpus with 145,179 records in this article. In this corpus, nine types of knowledge such as material, regulation method, product, faradaic efficiency, cell setup, electrolyte, synthesis method, current density, and voltage are provided by either annotating or extracting. Machine learning algorithms can be applied to the corpus to help scientists find new and effective electrocatalysts. Furthermore, researchers familiar with NLP can use this corpus to design domain-specific named entity recognition (NER) models.
RESUMO
We present a method of mapping data from publicly available genomics and publication resources to the Resource Description Framework (RDF) and implement a server to publish linked open data (LOD). As one of the largest and most comprehensive semantic databases about coronaviruses, the resulted gcCov database demonstrates the capability of using data in the LOD framework to promote correlations between genotypes and phenotypes. These correlations will be helpful for future research on fundamental viral mechanisms and drug and vaccine designs. These LOD with 62,168,127 semantic triplets and their visualizations are freely accessible through gcCov at https://nmdc.cn/gccov/.
RESUMO
BACKGROUND: Foodborne diseases, as a type of disease with a high global incidence, place a heavy burden on public health and social economy. Foodborne pathogens, as the main factor of foodborne diseases, play an important role in the treatment and prevention of foodborne diseases; however, foodborne diseases caused by different pathogens lack specificity in clinical features, and there is a low proportion of clinically actual pathogen detection in real life. OBJECTIVE: We aimed to analyze foodborne disease case data, select appropriate features based on analysis results, and use machine learning methods to classify foodborne disease pathogens to predict foodborne disease pathogens that have not been tested. METHODS: We extracted features such as space, time, and exposed food from foodborne disease case data and analyzed the relationship between these features and the foodborne disease pathogens using a variety of machine learning methods to classify foodborne disease pathogens. We compared the results of 4 models to obtain the pathogen prediction model with the highest accuracy. RESULTS: The gradient boost decision tree model obtained the highest accuracy, with accuracy approaching 69% in identifying 4 pathogens including Salmonella, Norovirus, Escherichia coli, and Vibrio parahaemolyticus. By evaluating the importance of features such as time of illness, geographical longitude and latitude, and diarrhea frequency, we found that they play important roles in classifying the foodborne disease pathogens. CONCLUSIONS: Data analysis can reflect the distribution of some features of foodborne diseases and the relationship among the features. The classification of pathogens based on the analysis results and machine learning methods can provide beneficial support for clinical auxiliary diagnosis and treatment of foodborne diseases.
RESUMO
BACKGROUND: Foodborne disease is a common threat to human health worldwide, leading to millions of deaths every year. Thus, the accurate prediction foodborne disease risk is very urgent and of great importance for public health management. OBJECTIVE: We aimed to design a spatial-temporal risk prediction model suitable for predicting foodborne disease risks in various regions, to provide guidance for the prevention and control of foodborne diseases. METHODS: We designed a novel end-to-end framework to predict foodborne disease risk by using a multigraph structural long short-term memory neural network, which can utilize an encoder-decoder to achieve multistep prediction. In particular, to capture multiple spatial correlations, we divided regions by administrative area and constructed adjacent graphs with metrics that included region proximity, historical data similarity, regional function similarity, and exposure food similarity. We also integrated an attention mechanism in both spatial and temporal dimensions, as well as external factors, to refine prediction accuracy. We validated our model with a long-term real-world foodborne disease data set, comprising data from 2015 to 2019 from multiple provinces in China. RESULTS: Our model can achieve F1 scores of 0.822, 0.679, 0.709, and 0.720 for single-month forecasts for the provinces of Beijing, Zhejiang, Shanxi and Hebei, respectively, and the highest F1 score was 20% higher than the best results of the other models. The experimental results clearly demonstrated that our approach can outperform other state-of-the-art models, with a margin. CONCLUSIONS: The spatial-temporal risk prediction model can take into account the spatial-temporal characteristics of foodborne disease data and accurately determine future disease spatial-temporal risks, thereby providing support for the prevention and risk assessment of foodborne disease.
RESUMO
Equitable and efficient allocation of pollutant discharge permits is vital for controlling total pollutant amounts. However, the conventional water pollutant discharge permit allocation method is criticized for dividing the environmental attributes of water bodies, which is mainly based on administrative units. China is establishing a water ecological environment zoning management system to manage the water environment more scientifically, which may have a great impact on for controlling total pollutant amounts. Whether the ecological environment zoning management system can promote more equitable and efficient permit allocation remains unknown. In this paper, an environmental zoning system and "basin-region" correlation are established to take both regional and watershed allocation processes into consideration. Then, a multi-index Gini coefficient method is established to evaluate the equality of different allocating methods. The Gini coefficient is then combined with a linear interactive and general optimizer method to achieve an equitable allocation of ammonia nitrogen discharge permits in the Songhua River Basin from both watershed and regional perspectives. Forty-five water pollutant discharge allocation scenarios are considered to represent different manager tendencies. The results show that allocation based on watershed functional units is more equitable than that based on administrative units. The index weighting settings also have a large impact on regional and total equality and environmental efficiency. Midstream and downstream areas show large allocation differences, although no scenario can satisfy all watershed regions in terms of equality and environmental efficiency at the same time. Thus, more trade-offs are needed during decision making. By considering the coordination of social, environmental and economic development at the basin level, this study provides new insight into equitable and efficient allocation. Moreover, the findings suggest that an environmental zoning system should be considered for discharge permit allocation in water resource management.
RESUMO
Building small-molecule libraries with structural and stereogenic diversity plays an important role in drug discovery. The development of switchable intermolecular cycloaddition reactions from identical substrates in different regioselective fashions would provide an attractive protocol. However, this also represents a challenge in organic chemistry, because it is difficult to control regioselectivity to afford the products exclusively and at the same time achieve high levels of stereoselectivity. Here, we report the diversified cycloadditions of α'-alkylidene-2-cyclopentenones catalysed by cinchona-derived primary amines. An asymmetric γ,ß'-regioselective intermolecular [6+2] cycloaddition reaction with 3-olefinic (7-aza)oxindoles is realized through the in situ generation of formal 4-aminofulvenes, while a different ß,γ-regioselective [2+2] cycloaddition reaction with maleimides to access fused cyclobutanes is disclosed. In contrast, an intriguing α,γ-regioselective [4+2] cycloaddition reaction is uncovered with the same set of substrates, by employing an unprecedented dual small-molecule catalysis of amines and thiols. All of the cycloaddition reactions exhibit excellent regio- and stereoselectivity, producing a broad spectrum of chiral architectures with high structural diversity and molecular complexity.