Integrated multiplex network based approach for hub gene identification in oral cancer

Abstract

Background:The incidence of Oral Cancer (OC) is high in Asian countries, which goes undetected at its early stage. The study of genetics, especially genetic networks holds great promise in this endeavor. Hub genes in a genetic network are prominent in regulating the whole network structure of genes. Thus identification of such genes related to specific cancer types can help in reducing the gap in OC prognosis. Methods: Traditional study of network biology is unable to decipher the inter-dependencies within and across diverse biological networks. Multiplex network provides a powerful representation of such systems and encodes much richer information than isolated networks. In this work, we focused on the entire multiplex structure of the genetic network integrating the gene expression profile and DNA methylation profile for OC. Further, hub genes were identified by considering their connectivity in the multiplex structure and the respective protein-protein interaction (PPI) network as well. Results: 46 hub genes were inferred in our approach with a high prediction accuracy (96%), outstanding Matthews coefficient correlation value (93%) and significant biological implications. Among them, genes PIK3CG, PIK3R5, MYH7, CDC20 and CCL4 were differentially expressed and predominantly enriched in molecular cascades specific to OC. Conclusions: The identified hub genes in this work carry ontological signatures specific to cancer, which may further facilitate improved understanding of the tumorigenesis process and the underlying molecular events. Result indicates the effectiveness of our integrated multiplex network approach for hub gene identification. This work puts an innovative research route for multi-omics biological data analysis.

Extra Tree Classifier Predicts an Interactome Hub Gene as HSPB1 in Oral Cancer: A Bioinformatics Analysis

“…One recent study identified 46 hub genes with 96% prediction accuracy, particularly PIK3CG, PIK3R5, MYH7, CDC20, and CCL4, which have significant biological implications for oral cancer and offered new research routes [21]. A study on head and neck squamous cell carcinoma patients identified 65 concordant genes, including a 13-gene panel.…”

Current advances in noninvasive methods for the diagnosis of oral squamous cell carcinoma: a review

“…Salivary biomarkers can be viewed from several perspectives, such as genomics, proteomics, metabolomics, and microbiomics (Fig. 2) Additionally, focus should be placed on comprehensive analysis of the integrative multi-omics module, where the multiplex network represents such systems effectively and encodes far more information than isolated networks [31,32].…”

PLEX.I: a tool to discover features in multiplex networks that reflect clinical variation

“…Each edge set, referred to as a layer, can represent a distinct type of data (e.g., transcriptomics and proteomics) or clinical condition (e.g., disease/health and responder/non-responder to a drug). Various methods have been proposed for the analysis of multiplex networks, many of which detect those nodes whose set of neighbors ( neighborhood ) is highly consistent across layers ( Buphamalai et al, 2021 ; Mahapatra et al, 2021 ; Peng et al, 2021 ). With highly consistent neighborhoods across layers may be prioritized as the most interesting ones for downstream computational or biological analysis.…”