Sparse Independence Component Analysis for Competitive Endogenous RNA Co-Module Identification in Liver Hepatocellular Carcinoma

Abstract

Objective: Long non-coding RNAs (lncRNAs) have been shown to be associated with the pathogenesis of different kinds of diseases and play important roles in various biological processes. Although numerous lncRNAs have been found, the functions of most lncRNAs and physiological/pathological significance are still in its infancy. Meanwhile, their expression patterns and regulation mechanisms are also far from being fully understood.

Methods: In order to reveal functional lncRNAs and identify the key lncRNAs, we develop a new sparse independence component analysis (ICA) method to identify lncRNA-mRNA-miRNA expression co-modules based on the competitive endogenous RNA (ceRNA) theory using the sample-matched lncRNA, mRNA and miRNA expression profiles. The expression data of the three RNA combined together is approximated sparsely to obtain the corresponding sparsity coefficient, and then it is decomposed by using ICA constraint optimization to obtain the common basis and modules. Subsequently, affine propagation clustering is used to perform cluster analysis on the common basis under multiple running conditions to obtain the co-modules for the selection of different RNA elements.

Results: We applied sparse ICA to Liver Hepatocellular Carcinoma (LIHC) dataset and the experiment results demonstrate that the proposed sparse ICA method can effectively discover biologically functional expression common modules.

Conclusion: It may provide insights into the function of lncRNAs and molecular mechanism of LIHC. Clinical and Translational Impact Statement-The results on LIHC dataset demonstrate that the proposed sparse ICA method can effectively discover biologically functional expression common modules, which may provide insights into the function of IncRNAs and molecular mechanism of LIHC.

Keywords: LIHC; Sparse ICA; ceRNA; co-expression modules; lncRNA.

FIGURE 1.

The flowchart of the sparse ICA method included different RNA data integration, sparse approximation coefficients formation and ICA decomposition and reconstruction, as well as co- module identification and functional analysis in LIHC.

FIGURE 2.

The evaluation results of sparse ICA performance. A) Histogram of sample-wise correlations of original and reconstructed miRNA, lncRNA and mRNA profiles across 424 samples, and the red line represents the standard deviation; B) Original data are plotted against the reconstructed miRNA, lncRNA and mRNA profiles with their correlation coefficients for three samples.

FIGURE 3.

Biological function enrichment of co- module. A) KEGG pathway enrichment analysis and GO enrichment analysis of DE mRNAs. The horizontal axis represents the gene ratio in the enriched pathway or GO biological process. Circle nodes and triangle nodes represent GO and KEGG pathway, respectively. The size of nodes denotes the number of genes in enrichment sets and the color of nodes denotes the significance of results; B) KEGG pathway enrichment analysis and GO enrichment analysis of DE coding genes regulated by DE miRNAs. Row represents the average number of genes regulated by each RNA. C) Disease enrichment analysis of DE mRNAs. The horizontal axis and the color of bar denote the significance of results. D) Disease enrichment analysis of mRNAs regulated by DE miRNAs. The horizontal axis and the color of bar denote the significance of results.

FIGURE 4.

A) CeRNA network constructed by using the co- module, the color of nodes represents the expression change of RNA compared to control samples, and Red (blue) indicates a significant increase (decrease). Round rectangle node represents miRNA, ellipse node denotes mRNA and diamond node represents lncRNA. The edge denotes the regulatory relationship between the two types of RNA; B) The association of lncRNAs and liver-related diseases from lncRNA disease database.

FIGURE 5.

The survival status of ceRNA interaction in LIHC. A) Forest plot of multivariable Cox regression analysis. The boxes on the transverse lines show the hazard ratio (HR), and the transverse lines represent 95% confidence interval (CI); B) Kaplan-Meier survival analysis for ceRNA interaction. CeRNA1 denotes the interaction of LINC00311, hsa-miR-218-5p and KLF9; CeRNA2 denotes the interaction of LINC00311, hsa-miR-222-3p and SOD2; CeRNA3 denotes the interaction of NEAT1, hsa-miR-222-3p and SOD2; CeRNA4 denotes the interaction of C5orf17, hsa-miR-222-3p and SOD2.