Deep Non-Negative Matrix Factorization Architecture Based on Underlying Basis Images Learning

Learning deep representation and discriminative features for clustering of multi-layer networks

The multi-layer network consists of the interactions between different layers, where each layer of the network is depicted as a graph, providing a comprehensive way to model the underlying complex systems. The layer-specific modules of multi-layer networks are critical to understanding the structure and function of the system. However, existing methods fail to characterize and balance the connectivity and specificity of layer-specific modules in networks because of the complicated inter- and intra-coupling of various layers. To address the above issues, a joint learning graph clustering algorithm (DRDF) for detecting layer-specific modules in multi-layer networks is proposed, which simultaneously learns the deep representation and discriminative features. Specifically, DRDF learns the deep representation with deep nonnegative matrix factorization, where the high-order topology of the multi-layer network is gradually and precisely characterized. Moreover, it addresses the specificity of modules with discriminative feature learning, where the intra-class compactness and inter-class separation of pseudo-labels of clusters are explored as self-supervised information, thereby providing a more accurate method to explicitly model the specificity of the multi-layer network. Finally, DRDF balances the connectivity and specificity of layer-specific modules with joint learning, where the overall objective of the graph clustering algorithm and optimization rules are derived. The experiments on ten multi-layer networks showed that DRDF not only outperforms eight baselines on graph clustering but also enhances the robustness of algorithms.

MCM2 is involved in subtyping and tamoxifen resistance of ERα-positive breast cancer by acting as the downstream factor of ERα

Tamoxifen (TAM) resistance is finally developed in over 40% of patients with estrogen receptor α-positive breast cancer (ERα+ -BC), documenting that discovering new molecular subtype is needed to confer perception to the heterogeneity of ERα+ -BC. We obtained representative gene sets subtyping ERα+ -BC using gene set variation analysis (GSVA), non-negative matrix factorization (NMF), and COX regression methods on the basis of METABRIC, TCGA, and GEO databases. Furthermore, the risk score of ERα+ -BC subtyping was established using least absolute shrinkage and selection operator (LASSO) regression on the basis of genes in the representative gene sets, thereby generating the two subtypes of ERα+ -BC. We further found that minichromosome maintenance complex component 2 (MCM2) functioned as the hub gene subtyping ERα+ -BC using GO, KEGG, and MCODE. MCM2 expression was capable for specifically predicting 1-year overall survival (OS) of ERα+ -BC and correlated with T stage, AJCC stage, and tamoxifen (TAM) sensitivity of ERα+ -BC. The downregulation of MCM2 expression inhibited proliferation, migration, and invasion of TAM-resistant cells and promoted G0/G1 arrest. Altogether, tamoxifen resistance entails that MCM2 is a hub gene subtyping ERα+ -BC, providing a novel dimension for discovering a potential target of TAM-resistant BC.

Hot Topic Recognition of Health Rumors Based on Anti-Rumor Articles on the WeChat Official Account Platform: Topic Modeling

BACKGROUND

Social networks have become one of the main channels for obtaining health information. However, they have also become a source of health-related misinformation, which seriously threatens the public's physical and mental health. Governance of health-related misinformation can be implemented through topic identification of rumors on social networks. However, little attention has been paid to studying the types and routes of dissemination of health rumors on the internet, especially rumors regarding health-related information in Chinese social media.

OBJECTIVE

This study aims to explore the types of health-related misinformation favored by WeChat public platform users and their prevalence trends and to analyze the modeling results of the text by using the Latent Dirichlet Allocation model.

METHODS

We used a web crawler tool to capture health rumor-dispelling articles on WeChat rumor-dispelling public accounts. We collected information from health-debunking articles posted between January 1, 2016, and August 31, 2022. Following word segmentation of the collected text, a document topic generation model called Latent Dirichlet Allocation was used to identify and generalize the most common topics. The proportion distribution of the themes was calculated, and the negative impact of various health rumors in different periods was analyzed. Additionally, the prevalence of health rumors was analyzed by the number of health rumors generated at each time point.

RESULTS

We collected 9366 rumor-refuting articles from January 1, 2016, to August 31, 2022, from WeChat official accounts. Through topic modeling, we divided the health rumors into 8 topics, that is, rumors on prevention and treatment of infectious diseases (1284/9366, 13.71%), disease therapy and its effects (1037/9366, 11.07%), food safety (1243/9366, 13.27%), cancer and its causes (946/9366, 10.10%), regimen and disease (1540/9366, 16.44%), transmission (914/9366, 9.76%), healthy diet (1068/9366, 11.40%), and nutrition and health (1334/9366, 14.24%). Furthermore, we summarized the 8 topics under 4 themes, that is, public health, disease, diet and health, and spread of rumors.

CONCLUSIONS

Our study shows that topic modeling can provide analysis and insights into health rumor governance. The rumor development trends showed that most rumors were on public health, disease, and diet and health problems. Governments still need to implement relevant and comprehensive rumor management strategies based on the rumors prevalent in their countries and formulate appropriate policies. Apart from regulating the content disseminated on social media platforms, the national quality of health education should also be improved. Governance of social networks should be clearly implemented, as these rapidly developed platforms come with privacy issues. Both disseminators and receivers of information should ensure a realistic attitude and disseminate health information correctly. In addition, we recommend that sentiment analysis-related studies be conducted to verify the impact of health rumor-related topics.

A Novel Clustering Method Based on Adjacent Grids Searching

Clustering is used to analyze the intrinsic structure of a dataset based on the similarity of datapoints. Its widespread use, from image segmentation to object recognition and information retrieval, requires great robustness in the clustering process. In this paper, a novel clustering method based on adjacent grid searching (CAGS) is proposed. The CAGS consists of two steps: a strategy based on adaptive grid-space construction and a clustering strategy based on adjacent grid searching. In the first step, a multidimensional grid space is constructed to provide a quantization structure of the input dataset. The noise and cluster halo are automatically distinguished according to grid density. Moreover, the adaptive grid generating process solves the common problem of grid clustering, in which the number of cells increases sharply with the dimension. In the second step, a two-stage traversal process is conducted to accomplish the cluster recognition. The cluster cores with arbitrary shapes can be found by concealing the halo points. As a result, the number of clusters will be easily identified by CAGS. Therefore, CAGS has the potential to be widely used for clustering datasets with different characteristics. We test the clustering performance of CAGS through six different types of datasets: dataset with noise, large-scale dataset, high-dimensional dataset, dataset with arbitrary shapes, dataset with large differences in density between classes, and dataset with high overlap between classes. Experimental results show that CAGS, which performed best on 10 out of 11 tests, outperforms the state-of-the-art clustering methods in all the above datasets.

Learning matrix factorization with scalable distance metric and regularizer

Matrix factorization has always been an encouraging field, which attempts to extract discriminative features from high-dimensional data. However, it suffers from negative generalization ability and high computational complexity when handling large-scale data. In this paper, we propose a learnable deep matrix factorization via the projected gradient descent method, which learns multi-layer low-rank factors from scalable metric distances and flexible regularizers. Accordingly, solving a constrained matrix factorization problem is equivalently transformed into training a neural network with an appropriate activation function induced from the projection onto a feasible set. Distinct from other neural networks, the proposed method activates the connected weights not just the hidden layers. As a result, it is proved that the proposed method can learn several existing well-known matrix factorizations, including singular value decomposition, convex, nonnegative and semi-nonnegative matrix factorizations. Finally, comprehensive experiments demonstrate the superiority of the proposed method against other state-of-the-arts.

MDSR-NMF: Multiple deconstruction single reconstruction deep neural network model for non-negative matrix factorization

Dimension reduction is one of the most sought-after strategies to cope with high-dimensional ever-expanding datasets. To address this, a novel deep-learning architecture has been designed with multiple deconstruction and single reconstruction layers for non-negative matrix factorization aimed at low-rank approximation. This design ensures that the reconstructed input matrix has a unique pair of factor matrices. The two-stage approach, namely, pretraining and stacking, aids in the robustness of the architecture. The sigmoid function has been adjusted in such a way that fulfils the non-negativity criteria and also helps to alleviate the data-loss problem. Xavier initialization technique aids in the solution of the exploding or vanishing gradient problem. The objective function involves regularizer that ensures the best possible approximation of the input matrix. The superior performance of MDSR-NMF, over six well-known dimension reduction methods, has been demonstrated extensively using five datasets for classification and clustering. Computational complexity and convergence analysis have also been presented to establish the model.

Application of non-negative matrix factorization in oncology: one approach for establishing precision medicine

The increase in the expectations of artificial intelligence (AI) technology has led to machine learning technology being actively used in the medical field. Non-negative matrix factorization (NMF) is a machine learning technique used for image analysis, speech recognition, and language processing; recently, it is being applied to medical research. Precision medicine, wherein important information is extracted from large-scale medical data to provide optimal medical care for every individual, is considered important in medical policies globally, and the application of machine learning techniques to this end is being handled in several ways. NMF is also introduced differently because of the characteristics of its algorithms. In this review, the importance of NMF in the field of medicine, with a focus on the field of oncology, is described by explaining the mathematical science of NMF and the characteristics of the algorithm, providing examples of how NMF can be used to establish precision medicine, and presenting the challenges of NMF. Finally, the direction regarding the effective use of NMF in the field of oncology is also discussed.

Dual-Graph Global and Local Concept Factorization for Data Clustering

Considering a wide range of applications of nonnegative matrix factorization (NMF), many NMF and their variants have been developed. Since previous NMF methods cannot fully describe complex inner global and local manifold structures of the data space and extract complex structural information, we propose a novel NMF method called dual-graph global and local concept factorization (DGLCF). To properly describe the inner manifold structure, DGLCF introduces the global and local structures of the data manifold and the geometric structure of the feature manifold into CF. The global manifold structure makes the model more discriminative, while the two local regularization terms simultaneously preserve the inherent geometry of data and features. Finally, we analyze convergence and the iterative update rules of DGLCF. We illustrate clustering performance by comparing it with latest algorithms on four real-world datasets.

Head and Neck Squamous Cell Carcinoma Subtypes Based on Immunologic and Hallmark Gene Sets in Tumor and Non-tumor Tissues

Background

Non-tumor tissue has a significant impact on the prognosis of head and neck squamous cell carcinoma (HNSCC). Previous studies for HNSCC have mainly focused on tumor tissue, greatly neglecting the role of non-tumor tissue. This study aimed to identify HNSCC subtypes and prognostic gene sets based on activity changes of immunologic and hallmark gene sets in tumor and adjacent non-tumor tissues to improve patient prognosis.

Methods

In the study, we used gene set variation analysis (GSVA) to estimate the relative enrichment of gene sets over the sample population, and identified relevant subtypes of HNSCC by Cox regression analysis and the non-negative matrix factorization (NMF) method. The representative gene sets were identified by calculating the differential enrichment score of gene sets between each of the two subgroups, intersecting them, and screening them using univariate Cox regression analysis. The least absolute shrinkage and selection operator (LASSO) regression analysis was used to screen out potential prognostic gene sets and establish a risk model. Finally, genes encompassed in each prognostic gene set were obtained and subjected to enrichment analysis and protein–protein interaction (PPI) in tumor and non-tumor tissues.

Results

We identified three subtypes of HNSCC based on gene sets in tumor and non-tumor tissues, and patients with subtype 1 had a higher survival rate than subtypes 2 and 3. The subtypes were related to the survival status, pathological stage, and T stage of HNSCC patients. In total 450 differentially gene sets and 39 representative gene sets were obtained by calculating the differential enrichment score of gene sets between each of the two subgroups, intersecting them, and screening them using univariate Cox regression analysis. The prognostic model was constructed by LASSO regression analysis, including five prognostic gene sets. Kaplan-Meier analysis indicated that different risk groups and the five prognostic gene sets were associated with survival status in the model. Finally, enrichment analysis and PPI indicated that non-tumor and tumor tissues affect the prognosis of HNSCC patients in different ways.

Conclusion

In conclusion, we provide a novel insight for rational treatment strategies and precise prognostic assessments based on tumor and adjacent non-tumor tissues, suggesting that more emphasis should be placed on changes in adjacent non-tumor and tumor tissues, rather than just the tumor itself.

Identification of immune-related subtypes of colorectal cancer to improve antitumor immunotherapy

Immunotherapy involving immune checkpoint inhibitors (ICIs) for enhancing immune system activation is promising for tumor management. However, the patients' responses to ICIs are different. Here, we applied a non-negative matrix factorization algorithm to establish a robust immune molecular classification system for colorectal cancer (CRC). We obtained data of 1503 CRC patients (training cohort: 488 from The Cancer Genome Atlas; validation cohort: 1015 from the Gene Expression Omnibus). In the training cohort, 42.8% of patients who exhibited significantly higher immunocyte infiltration and enrichment of immune response-associated signatures were subdivided into immune classes. Within the immune class, 53.1% of patients were associated with a worse overall prognosis and belonged to the immune-suppressed subclass, characterized by the activation of stroma-related signatures, genes, immune-suppressive cells, and signaling. The remaining immune class patients belonged to the immune-activated subclass, which was associated with a better prognosis and response to anti-PD-1 therapy. Immune-related subtypes were associated with different copy number alterations, tumor-infiltrating lymphocyte enrichment, PD-1/PD-L1 expression, mutation landscape, and cancer stemness. These results were validated in patients with microsatellite instable CRC. We described a novel immune-related class of CRC, which may be used for selecting candidate patients with CRC for immunotherapy and tailoring optimal immunotherapeutic treatment.