Robust hypergraph regularized non-negative matrix factorization for sample clustering and feature selection in multi-view gene expression data

Abstract

BackgroundAs one of the most popular data representation methods, non-negative matrix decomposition (NMF) has been widely concerned in the tasks of clustering and feature selection. However, most of the previously proposed NMF-based methods do not adequately explore the hidden geometrical structure in the data. At the same time, noise and outliers are inevitably present in the data.ResultsTo alleviate these problems, we present a novel NMF framework named robust hypergraph regularized non-negative matrix factorization (RHNMF). In particular, the hypergraph Laplacian regularization is imposed to capture the geometric information of original data. Unlike graph Laplacian regularization which captures the relationship between pairwise sample points, it captures the high-order relationship among more sample points. Moreover, the robustness of the RHNMF is enhanced by using the L2,1-norm constraint when estimating the residual. This is because the L2,1-norm is insensitive to noise and outliers.ConclusionsClustering and common abnormal expression gene (com-abnormal expression gene) selection are conducted to test the validity of the RHNMF model. Extensive experimental results on multi-view datasets reveal that our proposed model outperforms other state-of-the-art methods.

Modularization Design for Smart Industrial Service Ecosystem: A Framework Based on the Smart Industrial Service Identification Blueprint and Hypergraph Clustering

“…Hypergraph learning can be seen as a label propagation process or as a spectral partition in different tasks [60]. Hypergraph clustering or regularization is considered to be more robust to noise and interference than traditional graphs as it can provide higher-order relationships [42]. The advantages of hypergraph clustering can be utilized if it is applied to the modular design of the smart service solution.…”

Preprocessing of Single Cell RNA Sequencing Data Using Correlated Clustering and Projection

“…One such method is ScRNA by non-negative and low-rank representation (SinLRR), which assumes that scRNA-seq has an inherently low rank and attempts to find the smallest rank matrix that captures the original data . Numerous non-negative matrix factorization (NMF) methods with different constraints have also been developed, where the low-dimensional representation of scRNA-seq is a linear combination of the original data and acts as meta-genes. − Single-cell interpretation via multikernel learning (SIMLR) utilizes multiple kernels to learn a cell–cell similarity metric that generalizes to different biological experiments and experimental procedures . In addition, more traditional approaches, such as principal component analysis (PCA) and its derivatives, , and visualization techniques, such as uniform manifold approximation and projection (UMAP) and t-distributed stochastic neighbor embedding (t-SNE), have been heavily utilized for scRNA-seq data.…”

Robust Graph Regularized NMF with Dissimilarity and Similarity Constraints for ScRNA-seq Data Clustering

“…Xiao et al proposed a graph regularized NMF method (GRNMF) to discover potential associations between miRNAs and diseases in heterogeneous omics data. To alleviate the influence of noises, Yu et al developed a robust hypergraph regularized NMF for feature selection and sample clustering in gene expression data.…”

“…Xiao et al 16 proposed a graph regularized NMF method (GRNMF) to discover potential associations between miRNAs and diseases in heterogeneous omics data. To alleviate the influence of noises, Yu et al 17 In practice, the cell type labeling methods (manual labeling plus annotation library, immunofluorescence imaging, and cross-species comparison) can be used to label a small amount of data. In addition, there are some types of cells in clinical data, which can be used as labeled samples.…”