Scholarlink

<p>This study aims to refine molecular subtypes via multi-omics data, develop a prognostic signature, and identify novel biomarkers in colorectal cancer (CRC). On the basis of the multi-omics data, the <i>MOVICS</i> R package was used to divide patients with CRC into three consensus subtypes (CSs). Among the 3 CSs, CS1 had higher immune scores, immune checkpoint expression, infiltration of immune cells, and sensitivity to chemotherapy drugs. Specific markers of CS1 were identified, and 101 combinations of machine learning methods were applied for calculating the consensus machine learning-based score (CMLS) and constructing the CMLS signature for predicting patient survival. CMLS showed high efficiency in predicting the outcome of patients across multiple CRC datasets, and the results demonstrated that CMLS was an independent prognostic factor in CRC. The high- and low-CMLS groups presented distinct immune landscapes. CMLS was linked to malignant cancer features, suggesting its potential as a predictor of malignant progression in diverse cancers. Among the genes used to calculate CMLS, the role of inhibin subunit beta B (INHBB) in CRC remains unexplored. Expression analysis of INHBB across different cancer types revealed its upregulation in CRC, which was further validated by western blot and immunohistochemistry (IHC) experiments. INHBB silencing significantly inhibited tumor cell proliferation and migration, decreased phosphorylated AKT and N-cadherin levels, and increased E-cadherin expression. INHBB potentially suppresses CRC development and progression by suppressing the AKT signaling pathway and the EMT process.</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

Multi-omics based consensus subtypes, development of prognostic signature, and identification of INHBB as a potential therapeutic target in colorectal cancer

  • 期刊论文

  • 发表日期：

    2025年9月26日

查看全文
  • Xu Wang,
  • Yuanmin Xu,
  • Rui Sun,
  • Sheng Wang,
  • Xiang Wei

摘要

This study aims to refine molecular subtypes via multi-omics data, develop a prognostic signature, and identify novel biomarkers in colorectal cancer (CRC). On the basis of the multi-omics data, the MOVICS R package was used to divide patients with CRC into three consensus subtypes (CSs). Among the 3 CSs, CS1 had higher immune scores, immune checkpoint expression, infiltration of immune cells, and sensitivity to chemotherapy drugs. Specific markers of CS1 were identified, and 101 combinations of machine learning methods were applied for calculating the consensus machine learning-based score (CMLS) and constructing the CMLS signature for predicting patient survival. CMLS showed high efficiency in predicting the outcome of patients across multiple CRC datasets, and the results demonstrated that CMLS was an independent prognostic factor in CRC. The high- and low-CMLS groups presented distinct immune landscapes. CMLS was linked to malignant cancer features, suggesting its potential as a predictor of malignant progression in diverse cancers. Among the genes used to calculate CMLS, the role of inhibin subunit beta B (INHBB) in CRC remains unexplored. Expression analysis of INHBB across different cancer types revealed its upregulation in CRC, which was further validated by western blot and immunohistochemistry (IHC) experiments. INHBB silencing significantly inhibited tumor cell proliferation and migration, decreased phosphorylated AKT and N-cadherin levels, and increased E-cadherin expression. INHBB potentially suppresses CRC development and progression by suppressing the AKT signaling pathway and the EMT process.