Mining of Gene Modules and Identification of Key Genes for Early Diagnosis of Gastric Cancer

Ling Xu, Jinyan Yang, Yu Zhang, Xudong Liu, Zhe Liu, Feipeng Sun, Ying Ma, Lei Wang, Feng XingInternational Journal Of Pharmaceutical Research And Allied Sciences 2024 · DOI

TL;DR

This study leverages Weighted Gene Co-expression Network Analysis (WGCNA) and TCGA data to identify 19 key genes that show promise as early diagnostic biomarkers for gastric cancer (GC). By pinpointing these genes, the research aims to improve early detection, a critical factor in improving patient survival rates for this highly prevalent and deadly cancer.

What Problem Does This Paper Address?

Gastric cancer (GC) is a significant global health concern, characterized by high incidence and mortality rates. A major challenge in combating GC is its often late diagnosis. Many patients are identified only when the cancer has progressed to advanced stages, or during routine screenings, by which time the optimal window for effective treatment has often passed. This leads to lower survival rates and limited treatment options. The paper addresses the critical need for reliable early diagnostic biomarkers to enable timely intervention and improve patient outcomes.

Methods

The study employed a multi-pronged bioinformatics approach:

  • Differentially Expressed Gene (DEG) Screening: The "limma" R package was used to identify genes with significant expression differences between GC tumor samples and normal tissues from the TCGA database.
  • Functional Enrichment Analysis: Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed on the identified DEGs to understand their biological functions and associated pathways.
  • Weighted Gene Co-expression Network Analysis (WGCNA): This core method was used to construct a gene co-expression network. WGCNA clusters genes with similar expression patterns into modules and then correlates these modules with clinical traits (like disease status, gender, pathological stage, etc.). This helps identify modules that are strongly associated with GC.
  • Hub Gene Identification and Validation: Within the identified key modules (turquoise and blue, which showed significant correlation with GC), "hub genes" (genes central to the module's function) were selected based on module membership (MM) and gene significance (GS). These candidate hub genes were then further validated using survival analysis and differential expression analysis (via the GEPIA database) to assess their impact on patient prognosis and their expression differences between GC and normal tissues.
  • Diagnostic Value Analysis: Finally, one-way ANOVA was used to analyze the expression levels of the validated hub genes across different pathological stages of GC in normal tissues, aiming to assess their potential for early diagnosis.

Datasets

  • The Cancer Genome Atlas (TCGA) database: This was the primary source for RNA sequencing data and clinical information for GC patients. The dataset included 380 GC samples and 37 normal samples.

Key Insights and Ideas

  • WGCNA's Power in Uncovering Disease-Specific Modules: The study effectively demonstrates how WGCNA can identify specific gene modules (turquoise and blue) that are significantly correlated with GC, suggesting their involvement in the disease's pathogenesis or suppression.
  • Identification of Potential Tumorigenic and Anti-Tumor Modules: The turquoise module was found to be positively correlated with GC (suggesting a role in tumorigenesis), while the blue module showed a negative correlation (implying potential anti-tumor effects). This distinction provides valuable insights into the complex molecular landscape of GC.
  • Hub Genes as Potential Early Diagnostic Markers: The identification of 19 hub genes that not only correlate with GC but also show differential expression between normal and early-stage GC tissues is a significant finding. This suggests their potential utility in detecting the disease at its nascent stages.
  • Validation Across Multiple Platforms: The use of both TCGA data for initial screening and the GEPIA database for survival and expression validation strengthens the reliability of the identified hub genes.

Key Results

  • 4892 Differentially Expressed Genes (DEGs): The initial screening identified a substantial number of DEGs between GC and normal samples, comprising 3043 upregulated and 1849 downregulated genes.
  • Six Gene Modules Identified: WGCNA analysis revealed six distinct gene modules.
  • Two Clinically Significant Modules: The turquoise module showed a strong positive correlation (cor = 0.93) with GC, while the blue module showed a significant negative correlation (cor = -0.62).
  • 19 Verified Hub Genes: After rigorous screening and validation, 19 genes were identified as key hub genes associated with GC prognosis and expression.
  • Differential Expression in Early GC: Crucially, the study found that the expression levels of several of these hub genes (e.g., ASF1B, EPCAM, FAM83H, CEP55) were significantly different in early GC tissues compared to normal tissues, highlighting their potential as early diagnostic indicators.

Limitations and Future Work

  • Need for Experimental Validation: The authors acknowledge that their study is primarily based on bioinformatics analysis. They explicitly state that further molecular biology experiments are needed to confirm the precise functions of these hub genes in GC.
  • Shortcomings in Early Detection: While the study advances understanding of early GC diagnosis, it also points out that current methods still have limitations, and further research is needed to improve early detection rates.
  • Potential Therapeutic Targets: The identified hub genes, particularly those with significant roles in GC progression, could serve as potential therapeutic targets for future drug development.
  • Combination Biomarkers: The paper suggests that a combination of these hub genes, especially a subset like ASF1B, CEP55, PDZD4, ATP1A2, ABCA9, C8orf46, and TCEAL2, might be more effective as biomarkers for early GC detection.

Summary updated

Abstract

Gastric cancer (GC) is one of the most common malignant tumors with high incidence and mortality rates. Most patients with GC are not diagnosed until the advanced stage of cancer or during tumor screening, resulting in missing the best time for treatment. This study identified key modules and hub genes associated with GC by weighted gene co-expression network analysis (WGCNA). The "limma" package in R was used to identify differentially expressed genes (DEGs) in GC samples from TCGA, and a total
BibTeX
@article{xu2024amp,
  author = {Ling Xu and Jinyan Yang and Yu Zhang and Xudong Liu and Zhe Liu and Feipeng Sun and Ying Ma and Lei Wang and Feng Xing},
  title = {{\ Mining of Gene Modules and Identification of Key Genes for Early Diagnosis of Gastric Cancer}},
  journal = {International Journal Of Pharmaceutical Research And Allied Sciences},
  year = {2024},
  doi = {10.51847/mfoqlj1g2f},
}
Figure