Scholarlink

Background <p>Gastric cancer (GC) is a common digestive tract cancer whose high heterogeneity and invasiveness lead to a low survival rate. Therefore, it is necessary to explore the potential molecular mechanism of GC.</p> Methods <p>Three genes and one miRNA expression microarray dataset were downloaded from the GEO database. The gene expression profiles of the cancer group and normal group were compared in each dataset, and differentially expressed genes (DEGs) and miRNAs were identified with GEO2R. GO enrichment and KEGG pathway analysis of DEGs were performed with the R package clusterProfiler. The interaction network of DEGs was visualized with Cytoscape, and clusters and key genes were identified. According to the key DEGs, a prognostic risk model of GC was established by Cox regression. The patients were subdivided into high-risk and low-risk groups based on the median value of the risk score, and the model performance was evaluated by ROC curve and survival analyses. The risk model was combined with clinicopathological features to establish a nomogram, and a calibration chart was used to evaluate the prediction accuracy of the nomogram. The ROC curve was drawn to predict the ability of genes to differentiate tumour tissues from normal tissues. The starBase database was used to construct a regulatory network consistent with the miRNA–mRNA hypothesis. The expression of COL1A2 was analysed by immunohistochemistry/immunocytochemistry (IHC/ICC) in 150 patients with gastric cancer and a GC cell line. The correlations between COL1A2 expression and clinical features were analysed.</p> Results <p>A total of 106 DEGs and 113 differentially expressed miRNAs were identified from the gastric cancer dataset. PPI network screening revealed the two most significant modules and identified the dominant gene. A prognostic risk model composed of six prognostic genes (COL1A2, COL1A1, COL3A1, SPARC, LUM and BGN) was constructed by Cox regression analysis. The prognosis of the high-risk group was poor (<i>P</i> &lt; 0.001). ROC curve analysis (AUC = 0.732) showed that the model better predicted the 5-year survival rate of patients. In addition, the six prognostic genes had appropriate diagnostic ability for GC. A potential ceRNA network of model genes in GC was constructed through the starBase database. IHC showed that COL1A2 was highly expressed in gastric cancer and GC cells. COL1A2 expression was significantly correlated with lymph node metastasis.</p> Conclusion <p>This study reveals the potential biomarkers and related pathways of GC and provides a theoretical basis for the diagnosis and prognosis prediction of GC.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Collagen type I alpha 2 acts as a potential diagnostic biomarker and therapeutic targets for the prognosis in gastric cancer

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月31日

查看全文
  • Jingjing Dai,
  • Xudong Song,
  • Abdusemer Reyimu,
  • Zihao Gao,
  • Chenggong Zhang,
  • Fan Ding,
  • Dezhu Dai,
  • Liang Shi,
  • Xiao Han,
  • Wubi Zhou,
  • Guoquan Tao

摘要

Background

Gastric cancer (GC) is a common digestive tract cancer whose high heterogeneity and invasiveness lead to a low survival rate. Therefore, it is necessary to explore the potential molecular mechanism of GC.

Methods

Three genes and one miRNA expression microarray dataset were downloaded from the GEO database. The gene expression profiles of the cancer group and normal group were compared in each dataset, and differentially expressed genes (DEGs) and miRNAs were identified with GEO2R. GO enrichment and KEGG pathway analysis of DEGs were performed with the R package clusterProfiler. The interaction network of DEGs was visualized with Cytoscape, and clusters and key genes were identified. According to the key DEGs, a prognostic risk model of GC was established by Cox regression. The patients were subdivided into high-risk and low-risk groups based on the median value of the risk score, and the model performance was evaluated by ROC curve and survival analyses. The risk model was combined with clinicopathological features to establish a nomogram, and a calibration chart was used to evaluate the prediction accuracy of the nomogram. The ROC curve was drawn to predict the ability of genes to differentiate tumour tissues from normal tissues. The starBase database was used to construct a regulatory network consistent with the miRNA–mRNA hypothesis. The expression of COL1A2 was analysed by immunohistochemistry/immunocytochemistry (IHC/ICC) in 150 patients with gastric cancer and a GC cell line. The correlations between COL1A2 expression and clinical features were analysed.

Results

A total of 106 DEGs and 113 differentially expressed miRNAs were identified from the gastric cancer dataset. PPI network screening revealed the two most significant modules and identified the dominant gene. A prognostic risk model composed of six prognostic genes (COL1A2, COL1A1, COL3A1, SPARC, LUM and BGN) was constructed by Cox regression analysis. The prognosis of the high-risk group was poor (P < 0.001). ROC curve analysis (AUC = 0.732) showed that the model better predicted the 5-year survival rate of patients. In addition, the six prognostic genes had appropriate diagnostic ability for GC. A potential ceRNA network of model genes in GC was constructed through the starBase database. IHC showed that COL1A2 was highly expressed in gastric cancer and GC cells. COL1A2 expression was significantly correlated with lymph node metastasis.

Conclusion

This study reveals the potential biomarkers and related pathways of GC and provides a theoretical basis for the diagnosis and prognosis prediction of GC.