Poor Clusters Research Articles

Uncertain data density peak clustering algorithm based on JS divergence

Aiming at the defects of traditional density-based uncertainty clustering algorithms, such as parameter sensitivity and poor clustering results for complex manifold uncertain data sets, a new uncertainty data density peak clustering algorithm based on JS divergence (UDPC-JS) is proposed. The algorithm first removes noise points by using the uncertain natural neighbor density factor defined by the uncertain natural neighbor. Secondly, the local density of uncertain data objects is calculated by combining the uncertain natural neighbor and JS divergence. The initial clustering center of the uncertain data set is found by combining the idea of representative points, and the distance between the initial clustering centers is defined based on JS divergence and graph. Then, the decision graph is constructed on the initial clustering center using the local density calculated based on the uncertain natural neighbor and JS divergence and the newly defined distance between the initial clustering centers based on JS divergence and graph, and the final clustering center is selected according to the decision graph. Finally, the unassigned uncertain data objects are assigned to the cluster where their initial clustering center is located. Experimental results show that the algorithm has better clustering effect and accuracy than the comparison algorithm, and has a greater advantage in processing complex manifold uncertain data sets.

Application of Improved HF-DBSCAN Algorithm in Analyzing Complex Trajectory Data of Wi-Fi Users in Smart Campus

ABSTRACT With the development of wireless network technology and the increasingly widespread application of mobile intelligent terminals, a large amount of wireless network data has been generated in smart campuses. By analyzing the trajectory data through clustering algorithms, more valuable user data information can be obtained. However, current clustering algorithms have high computational difficulty, long running time, and poor clustering accuracy for complex trajectories. Based on this background, the study proposes a spatial clustering of Application with Noise by Density-Based Spatial Clustering of Application with Noise combining Hausdorff and Frechet. The method is denoted as (Hausdorff and Frechet-Density Based Spatial Clustering of Application with Noise, HF-DBSCAN). DBSCAN denotes the Density Based Spatial Clustering of Application with Noise. In the experimental results, compared to the AH-DBSCAN, DBSCAN, and FD-DBSCAN algorithms, the running time of HF-DBSCAN algorithm in complex trajectory cluster analysis is reduced by 59.33%, 39.82%, and 35.12%, respectively, and the contour coefficient is closer to 1; the Davies-Boldin Index (DBI) values decreased by 59.24%, 64.42%, and 68.24%, respectively. The experiment shows that the optimized HF-DBSCAN algorithm has lower computational difficulty, better clustering performance, and higher clustering accuracy, which verifies the effectiveness of this study.

Cancer subtype identification by multi-omics clustering based on interpretable feature and latent subspace learning

In recent years, multi-omics clustering has become a powerful tool in cancer research, offering a comprehensive perspective on the diverse molecular characteristics inherent to various cancer subtypes. However, most existing multi-omics clustering methods directly integrate heterogeneous features from different omics, which may struggle to deal with the noise or redundancy of multi-omics data and lead to poor clustering results. Therefore, we propose a novel multi-omics clustering method to extract interpretable and discriminative features from various omics before data integration. The clinical information is used to supervise the process of feature extraction based on SHAP (SHapley Additive exPlanation) values. Singular value decomposition (SVD) is then applied to integrate the extracted features of different omics by constructing a latent subspace. Finally, we utilize shared nearest neighbor-based spectral clustering on the latent representation to obtain the clustering result. The proposed method is evaluated on several cancer datasets across three levels of omics, in comparison to several state-of-the-art multi-omics clustering methods. The comparison results demonstrate the superior performance of the proposed method in multi-omics data analysis for cancer subtyping. Additionally, experiments reveal the efficacy of utilizing clinical information based on SHAP values for feature extraction, enhancing the performance of clustering analyses. Moreover, enrichment analysis of the identified gene signatures in different subtypes is also performed to further demonstrate the effectiveness of the proposed method.Availability: The proposed method can be freely accessible at https://github.com/Tianyi-Shi-Tsukuba/Multi-omics-clustering-based-on-SHAP. Data will be made available on request.

A Novel Classification Model Based on Hybrid K-Means and Neural Network for Classification Problems

We propose a new classification model—a new classification model for clustering overlapping problems based on K-Means and neural networks. K-means clustering algorithm belongs to unsupervised learning. It is a classic algorithm for solving clustering problems. Since this algorithm calculates its categories based on distance, the results tend to converge to the local optimal solution and have poor boundary clustering properties. The K-Means classification algorithm defines clusters by the distance between the cluster center value and the target object, and the optimal result is obtained through continuous iteration. Therefore, clustering results are overlapped, and there are often outliers that do not belong to the current cluster, resulting in unsatisfactory clustering results. Our model offers a new method to segment non-ideal data in overlapping regions. Since clustering algorithms cannot effectively identify and classify this part of the data, we split this part of the data and train it using a neural network. The results are then integrated into the clustered data. In the experiment, the k-fold cross-validation method ensures the model stability of the results. We used the accuracy to evaluate the quality of the model, and we used standard deviation and mean deviation to detect clustering results. Five sets of experimental data from the cross-experiment show that compared with the K-Means classification model, the accuracy of our model is effectively improved. Doi: 10.28991/HIJ-2024-05-03-012 Full Text: PDF

SPGMVC: Multiview Clustering via Partitioning the Signed Prototype Graph.

Multiview clustering (MVC) has been widely studied in machine learning and data mining for its capability of improving clustering performance by fusing the information from multiview data. In the past decade, a large number of MVC methods have made impressive progress, but most of them suffer from computational burdens, especially in large-scale tasks. Binary MVC (BMVC) is proposed to address this issue by representing the large-scale high-dimensional dataset as a group of consensus and low-dimensional binary codes. However, current BMVC-based approaches generate the clustering by executing binary k -means on the obtained binary codes, which fail to capture the embedded geometric information, leading to poor clustering performance. In addition, parameter selection is another "mission impossible" in unsupervised learning tasks including MVC. To tackle these challenges, a framework of multiview clustering via partitioning the signed prototype graph (SPGMVC) is proposed in this work. The SPGMVC framework offers several contributions. First, SPGMVC is designed as a unified framework for MVC. It combines effective technologies, such as consensus binary coding, code compression (CC), signed prototype graph (SPG) partitioning, and prototype-based cluster assignment. Second, SPGMVC partitions the signed graph (SG) based on the relationships between positive and negative edges. By capturing the underlying structure of the data, this partitioning strategy improves clustering accuracy (ACC). CC techniques are applied to reduce the graph's scale, enabling further partitioning and enhancing computational efficiency. Third, SPGMVC employs an alternate minimizing strategy to efficiently handle the optimization problem. This strategy has nearly linear time and space complexity with respect to the data volume, making it suitable for large-scale tasks. Fourth, SPGMVC proposes an automatic parameter selection strategy, eliminating the need for extensive parameter exploration. Comprehensive experiments illustrate the superiority of our model. The implementation of SPGMVC is available at: https://github.com/gepingyang/PSGMVC.

Patient-specific computational models predict prognosis in B cell lymphoma by quantifying pro-proliferative and anti-apoptotic signatures from genetic sequencing data

Genetic heterogeneity and co-occurring driver mutations impact clinical outcomes in blood cancers, but predicting the emergent effect of co-occurring mutations that impact multiple complex and interacting signalling networks is challenging. Here, we used mathematical models to predict the impact of co-occurring mutations on cellular signalling and cell fates in diffuse large B cell lymphoma and multiple myeloma. Simulations predicted adverse impact on clinical prognosis when combinations of mutations induced both anti-apoptotic (AA) and pro-proliferative (PP) signalling. We integrated patient-specific mutational profiles into personalised lymphoma models, and identified patients characterised by simultaneous upregulation of anti-apoptotic and pro-proliferative (AAPP) signalling in all genomic and cell-of-origin classifications (8-25% of patients). In a discovery cohort and two validation cohorts, patients with upregulation of neither, one (AA or PP), or both (AAPP) signalling states had good, intermediate and poor prognosis respectively. Combining AAPP signalling with genetic or clinical prognostic predictors reliably stratified patients into striking prognostic categories. AAPP patients in poor prognosis genetic clusters had 7.8 months median overall survival, while patients lacking both features had 90% overall survival at 120 months in a validation cohort. Personalised computational models enable identification of novel risk-stratified patient subgroups, providing a valuable tool for future risk-adapted clinical trials.

An adaptive kernel dictionary-based low-rank representation method for subspace clustering

Low-rank representation (LRR) is a classic subspace clustering (SC) algorithm, and many LRR-based methods have been proposed. Generally, LRR-based methods use denoized data as dictionaries for data reconstruction purpose. However, the dictionaries used in LRR-based algorithms are fixed, leading to poor clustering performance. In addition, most of these methods assume that the input data are linearly correlated. However, in practice, data are mostly nonlinearly correlated. To address these problems, we propose a novel adaptive kernel dictionary-based LRR (AKDLRR) method for SC. Specifically, to explore nonlinear information, the given data are mapped to the Hilbert space via the kernel technique. The dictionary in AKDLRR is not fixed; it adaptively learns from the data in the kernel space, making AKDLRR robust to noise and yielding good clustering performance. To solve the AKDLRR model, an efficient procedure including an alternative optimization strategy is proposed. In addition, a theoretical analysis of the convergence performance of AKDLRR is presented, which reveals that AKDLRR can converge in at most three iterations under certain conditions. The experimental results show that AKDLRR can achieve the best clustering performance and has excellent speed in comparison with other algorithms.

Habitat Enrichment Causes Changes in Fish Behavioural Characteristics: A Case Study of Sparus latus.

To better understand the habitat preferences and behavioural ecology of Sparus latus, we performed an experiment using box-shaped reefs as habitat enrichment materials, allowing us to determine the behavioural strategies and drivers involved in the response to different enrichment structures. The results showed that the first contact time of S. latus was negatively correlated (Pearson's correlation, p < 0.005) with the distribution rate in the artificial reef area. Enrichment structures affected the habitat preferences of S. latus, and there was a significant difference in the average distribution rate between the control and treatment groups (Adonis, p < 0.001). The opening ratio (Adonis, R2 = 0.36) explained the distribution difference of S. latus better than the opening shape (Adonis, R2 = 0.12). In the absence of an enrichment structure, S. latus remained more active during the daytime, exhibiting poor clustering, while in the presence of an enrichment structure, S. latus exhibited clustered movement at night. The opening ratio was negatively correlated with the average interindividual distance (Spearman's correlation, p < 0.01) and showed a significant positive correlation with the average distribution rate in the reef area (Spearman's correlation, p < 0.001), indicating that the reef opening ratio significantly affected the reef-tropism and clustering behaviours of S. latus. The light intensity was negatively correlated with the average distance moved, and the average speed (Spearman's correlation, p < 0.05) was significantly positively correlated with the reef first contact time (Spearman's correlation, p < 0.001), indicating that the light intensity affected the exploration and activity patterns of S. latus. These results provide a research basis for analysing the pattern and process of fish proliferation induced by artificial reef habitats.

Abstract 1155: Symptomatic Low- And High-degree Carotid Stenoses Exhibit Similar Features Of Plaque Instability

Background: Carotid endarterectomy (CEA) for symptomatic carotid stenosis is recommended for patients with the &gt;70% (NASCET) stenosis but not for those with &lt;50%. Since low-degree stenoses may still cause strokes, refined risk stratification is necessary, which could be improved by incorporating biological features of plaque instability. In order to challenge risk-stratification by the degree of stenosis, we compared biological features of carotid plaques from symptomatic patients with &lt;50% stenosis vs. &gt;70% and explored mechanisms behind plaque instability in low-degree stenoses. Methods: CEA specimens were retrieved from symptomatic patients with &gt;70% (n=204) and &lt;50% stenosis (n=34), all part of the Biobank of Karolinska Endarterectomies (BiKE). Patient demographics, plaque morphology from image analyses of preoperative computed tomography angiography (CTA), and RNA-seq analyses of lesions were used for comparisons. Plaque biology was assessed by transcriptomics using principal component analyses (PCA), differential gene expression (DGE)- and gene-set enrichment analyses (GSEA). Immunohistochemistry (IHC) was used to study proteins corresponding to upregulated genes. Results: The demographics of the two groups were statistically similar. They also showed similar plaque morphology with respect to intraplaque hemorrhage (IPH), plaque burden and fibrous cap thickness, whereas calcification was higher in the &lt;50% group and lipid-rich necrotic core higher in the &gt;70% group. PCA analysis indicated poor clustering, while DGE and GSEA identified enrichment of genes and pathways related to tissue hypoxia and angiogenesis in &lt;50% lesions. Corresponding proteins were detected by IHC in neo-vascularized plaque regions. Conclusions: Plaques from symptomatic patients with &lt;50% carotid stenoses exhibit largely similar morphological and biological features as those from &gt;70% stenoses, emphasizing the need for improved stratification of patients for CEA of symptomatic carotid stenosis. However, pathways associating plaque burden, hypoxia and angiogenesis were particularly enriched in low-grade stenosis lesions suggesting mechanisms behind plaque instability and symptoms in patients with both low- and high-degree stenosis.

Energy big data abnormal cluster detection method based on redundant convolution codec

Due to the scattered distribution and poor clustering of abnormal clusters in energy big data, the ability to detect anomalies is poor. Therefore, a high-energy data anomaly clustering detection method based on redundant convolutional encoding is proposed. Quantitative analysis of the coupling characteristics of electrical thermal gas optical time series for multi energy users based on Copula function, and incorporating quantitative values into multi energy feature indicators to extract the energy consumption behavior characteristics of multi energy users. Utilize redundant convolutional codecs to recombine and structurally encode abnormal features of energy big data, and capture multi energy coupling time features using coupling time capsule layers. Then, coupling time features are synthesized through fully connected linear regression layers to generate anomalous clustering feature components, and the energy time series data is then transformed into feature values of the time series in three-dimensional space. Based on this, a comprehensive energy system and massive multi energy user energy big data anomaly clustering analysis are carried out to determine the optimal number of multi energy users. Then, based on linear layers, the electricity heat gas light load characteristic map of multi energy users is transformed into one-dimensional form, and an energy big data anomaly clustering detection model is constructed to complete anomaly detection. The simulation results show that the proposed method has excellent feature clustering performance, detection accuracy above 98.7%, fast convergence speed, and an error rate below 0.1, which has reliable application value.

A novel tree structure-based multi-prototype clustering algorithm

Density-peak clustering (DPC) is a novel clustering algorithm that has received much attention from researchers. Its principle is to identify the cluster center of the data by density and distance. However, the density-peak clustering algorithm still suffers from poor clustering results due to irrational choices of cluster centers and misallocating samples. To tackle these problems, a novel tree structure-based multi-prototype clustering algorithm (MPCTS) is proposed in this paper. Firstly, a division-merger method is adopted to divide the dataset into multiple peak clusters based on multi-prototype clustering, which could minimize the impact of high-density sample points on the subsequent sample point assignments. Secondly, the KNN nearest-neighbor search strategy is improved to quickly obtain the sub-cluster center in each peak cluster to overcome the defect of irrational selection of initial cluster centers. Afterward, a tree-based decay strategy is designed to evaluate the correlation between sample points, and a tree-based association strategy is designed to achieve the merging of multiple peak clusters to tackle the sample points misallocation problem. Finally, the extensive experiments on clustering task are conducted over several benchmark synthetic and real datasets to verify the efficiency and superiority of the proposed MPCTS.

Trustworthy multi-view clustering via alternating generative adversarial representation learning and fusion

Multi-view clustering (MVC) has attached extensive attention as it provides an effective approach to deal with the unlabeled data in real-world applications. To enjoy the strong feature extraction capacity of deep learning, traditional shallow MVC methods are further extended to deep version. Though achieve superiorities in many fields, most existing deep MVC models are still limited by their lacking assurance of the trustworthiness and robustness. This drawback also leads to the possible poor clustering performance. To tackle this drawback, we propose an alternating generative adversarial learning strategy based network, termed as Alternating Generative Adversarial Representation Learning (AGARL) network for multi-view clustering, aiming to drive different views to fall into the same semantic space for the further alignment and information fusion. Various from the traditional generative adversarial learning strategy, the proposed alternating adversarial representation learning strategy successes in further aligning the distributions of the view-specific representations and exploring the cross-view consistency credited by its setting each view as the reference alternately. Then a consensus representative latent representation can be learned through an attention-like sub-network for all views. Finally, a self-supervised clustering module is employed in the proposed network to guide the learning of cluster assignment. In addition, comprehensive experiments on numerous benchmark MVC datasets have been performed to demonstrate the effectiveness of the proposed network compared to other state-of-the-art methods.

Influence of Optimization of the k-Means Algorithm with Genetic Algorithm on the Results of High Dimension Data Clustering

Clustering k-means begins with the random initial determination of the centroid. Initially generated random centroids often cause k-means to be trapped in the optimum local solution, which results in poor clustering quality. Therefore, this study examined the effect of genetic algorithms in determining initial centroids in k-means. Clustering k-means with random initial centroids and with initial centroids from genetic algorithm calculations are each tested on the data with dimension reduction and without dimension reduction. Based on the results of the initial centroid testing obtained from genetic algorithms, the quality of cluster results increased by 54.9% in the high dimensional data and 52.4% in the data that had been carried out for the dimensional reduction. This result shows that the k-means clustering with initial centroids obtained from genetic algorithm calculations has the best cluster/solution results with significant results.

DRBM-ClustNet: A Deep Restricted Boltzmann-Kohonen Architecture for Data Clustering.

A Bayesian deep restricted Boltzmann-Kohonen architecture for data clustering termed deep restricted Boltzmann machine (DRBM)-ClustNet is proposed. This core-clustering engine consists of a DRBM for processing unlabeled data by creating new features that are uncorrelated and have large variance with each other. Next, the number of clusters is predicted using the Bayesian information criterion (BIC), followed by a Kohonen network (KN)-based clustering layer. The processing of unlabeled data is done in three stages for efficient clustering of the nonlinearly separable datasets. In the first stage, DRBM performs nonlinear feature extraction by capturing the highly complex data representation by projecting the feature vectors of d dimensions into n dimensions. Most clustering algorithms require the number of clusters to be decided a priori; hence, here, to automate the number of clusters in the second stage, we use BIC. In the third stage, the number of clusters derived from BIC forms the input for the KN, which performs clustering of the feature-extracted data obtained from the DRBM. This method overcomes the general disadvantages of clustering algorithms, such as the prior specification of the number of clusters, convergence to local optima, and poor clustering accuracy on nonlinear datasets. In this research, we use two synthetic datasets, 15 benchmark datasets from the UCI Machine Learning repository, and four image datasets to analyze the DRBM-ClustNet. The proposed framework is evaluated based on clustering accuracy and ranked against other state-of-the-art clustering methods. The obtained results demonstrate that the DRBM-ClustNet outperforms state-of-the-art clustering algorithms.

The role of aldehydes on sulfur based-new particle formation: a theoretical study.

Aldehydes play a crucial role in the formation of atmospheric particles, attracting significant attention due to their environmental impact. However, the microscopic mechanisms underlying the formation of aldehyde-involved particles remain uncertain. In this study, through quantum chemical calculations and molecular dynamics (MD) simulations, we investigate the microscopic formation mechanisms of binary and ternary systems composed of three representative aldehydes, two sulfur-based acids, water, and two bases. Our research findings reveal that the most stable structures of acid-aldehyde clusters involve the connection of acids and aldehyde compounds through hydrogen bonds without involving proton transfer reactions, indicating relatively poor cluster stability. However, with the introduction of a third component, the stability of 18 clusters significantly increase. Among these, in ten systems, acids act as catalysts, facilitating reactions between aldehyde compounds and water or alkaline substances to generate glycols and amino alcohols. However, according to MD simulations conducted at 300 K, these acids readily dissociate from the resulting products. In the remaining eight systems, the most stable structural feature involves ion pairs formed by proton transfer reactions between acids and aldehyde compounds. These clusters exhibit remarkable thermodynamic stability. Furthermore, the acidity of the acid, the nature of nucleophilic agents, and the type of aldehyde all play significant roles in cluster stability and reactivity, and they have synergistic effects on the nucleation process. This study offers microscopic insights into the processes of new particle formation involving aldehydes, contributing to a deeper understanding of atmospheric chemistry at the molecular level.

Redundancy-Free Self-Supervised Relational Learning for Graph Clustering.

Graph clustering, which learns the node representations for effective cluster assignments, is a fundamental yet challenging task in data analysis and has received considerable attention accompanied by graph neural networks (GNNs) in recent years. However, most existing methods overlook the inherent relational information among the nonindependent and nonidentically distributed nodes in a graph. Due to the lack of exploration of relational attributes, the semantic information of the graph-structured data fails to be fully exploited which leads to poor clustering performance. In this article, we propose a novel self-supervised deep graph clustering method named relational redundancy-free graph clustering (R 2 FGC) to tackle the problem. It extracts the attribute-and structure-level relational information from both global and local views based on an autoencoder (AE) and a graph AE (GAE). To obtain effective representations of the semantic information, we preserve the consistent relationship among augmented nodes, whereas the redundant relationship is further reduced for learning discriminative embeddings. In addition, a simple yet valid strategy is used to alleviate the oversmoothing issue. Extensive experiments are performed on widely used benchmark datasets to validate the superiority of our R 2 FGC over state-of-the-art baselines. Our codes are available at https://github.com/yisiyu95/R2FGC.

Avoiding inferior clusterings with misspecified Gaussian mixture models

Clustering is a fundamental tool for exploratory data analysis, and is ubiquitous across scientific disciplines. Gaussian Mixture Model (GMM) is a popular probabilistic and interpretable model for clustering. In many practical settings, the true data distribution, which is unknown, may be non-Gaussian and may be contaminated by noise or outliers. In such cases, clustering may still be done with a misspecified GMM. However, this may lead to incorrect classification of the underlying subpopulations. In this paper, we identify and characterize the problem of inferior clustering solutions. Similar to well-known spurious solutions, these inferior solutions have high likelihood and poor cluster interpretation; however, they differ from spurious solutions in other characteristics, such as asymmetry in the fitted components. We theoretically analyze this asymmetry and its relation to misspecification. We propose a new penalty term that is designed to avoid both inferior and spurious solutions. Using this penalty term, we develop a new model selection criterion and a new GMM-based clustering algorithm, SIA. We empirically demonstrate that, in cases of misspecification, SIA avoids inferior solutions and outperforms previous GMM-based clustering methods.

A MapReduce-based approach to social network big data mining

The rapid development of social networks has facilitated the convenience of users to receive information. As a network communication platform for people’s daily use, microblog has countless information data. In view of the low efficiency and poor clustering effect of K-means algorithm, a parallel K-means clustering algorithm based on MapReduce model is studied; In order to alleviate the difficulty in calculating the similarity of microblog topic text, the space vector model and semantic similarity are used to calculate the similarity between texts to improve the quality of microblog text classification. The data expansion rate of corresponding nodes under different data sets shows that the average expansion rate of the parallel K-means algorithm reaches 0.89, and the running rate is the highest. The results show that the parallel K-means algorithm has good clustering stability and the highest clustering quality, reaching 1.24; The clustering time of the algorithm is the shortest, the average clustering time is 1.27 minutes, and the clustering effect and efficiency of the algorithm are the best. In the quality analysis of Weibo topic recommendation, the accuracy of P-K-means recommendation is 95.64%, user satisfaction is 98.64%, and the recommendation effect is also the best. It shows that the research on the parallel K-means clustering algorithm based on MapReduce has the best performance in microblogging topic mining and recommendation, which can efficiently recommend topics of interest to users and enhance users’ microblogging experience.

VERTICO. VII. Environmental Quenching Caused by the Suppression of Molecular Gas Content and Star Formation Efficiency in Virgo Cluster Galaxies

We study how environment regulates the star formation cycle of 33 Virgo Cluster satellite galaxies on 720 pc scales. We present the resolved star-forming main sequence for cluster galaxies, dividing the sample based on their global H i properties and comparing to a control sample of field galaxies. H i–poor cluster galaxies have reduced star formation rate (SFR) surface densities with respect to both H i–normal cluster and field galaxies (∼0.5 dex), suggesting that mechanisms regulating the global H i content are responsible for quenching local star formation. We demonstrate that the observed quenching in H i–poor galaxies is caused by environmental processes such as ram pressure stripping (RPS), simultaneously reducing the molecular gas surface density and star formation efficiency (SFE) compared to regions in H i–normal systems (by 0.38 and 0.22 dex, respectively). We observe systematically elevated SFRs that are driven by increased molecular gas surface densities at fixed stellar mass surface density in the outskirts of early stage RPS galaxies, while SFE remains unchanged with respect to the field sample. We quantify how RPS and starvation affect the star formation cycle of inner and outer galaxy disks as they are processed by the cluster. We show both are effective quenching mechanisms, with the key difference being that RPS acts upon the galaxy outskirts while starvation regulates the star formation cycle throughout disk, including within the truncation radius. For both processes, the quenching is caused by a simultaneous reduction in the molecular gas surface densities and SFE at fixed stellar mass surface density.

ConSpaS: a contrastive learning framework for identifying spatial domains by integrating local and global similarities.

Spatial transcriptomics is a rapidly growing field that aims to comprehensively characterize tissue organization and architecture at single-cell or sub-cellular resolution using spatial information. Such techniques provide a solid foundation for the mechanistic understanding of many biological processes in both health and disease that cannot be obtained using traditional technologies. Several methods have been proposed to decipher the spatial context of spots in tissue using spatial information. However, when spatial information and gene expression profiles are integrated, most methods only consider the local similarity of spatial information. As they do not consider the global semantic structure, spatial domain identification methods encounter poor or over-smoothed clusters. We developed ConSpaS, a novel node representation learning framework that precisely deciphers spatial domains by integrating local and global similarities based on graph autoencoder (GAE) and contrastive learning (CL). The GAE effectively integrates spatial information using local similarity and gene expression profiles, thereby ensuring that cluster assignment is spatially continuous. To improve the characterization of the global similarity of gene expression data, we adopt CL to consider the global semantic information. We propose an augmentation-free mechanism to construct global positive samples and use a semi-easy sampling strategy to define negative samples. We validated ConSpaS on multiple tissue types and technology platforms by comparing it with existing typical methods. The experimental results confirmed that ConSpaS effectively improved the identification accuracy of spatial domains with biologically meaningful spatial patterns, and denoised gene expression data while maintaining the spatial expression pattern. Furthermore, our proposed method better depicted the spatial trajectory by integrating local and global similarities.