Scholarlink

<p>Lung squamous cell carcinoma (LUSC) is characterized by high aggressiveness and significant heterogeneity, currently lacking highly precise individualized treatment options. Leveraging computational methodologies, we integrated multi-omics data from LUSC patients using 10 clustering algorithms, coupled with 10 machine learning algorithms, to delineate high-resolution molecular subtypes and develop a robust consensus machine learning-driven signature (CMLS). Through multi-omics clustering, we identified two multiomics subtypes (MoSs) associated with prognosis, with MoS1 exhibiting the most favorable prognostic outcome. MoS2 demonstrated a heightened immune-infiltrated state and a greater likelihood of responding to immunotherapy, while MoS1, although potentially less responsive to immunotherapy, may benefit more from Erlotinib. Moreover, we sought potential subchemotherapeutic targets based on the molecular expression characteristics of MoSs. Subsequent screening enabled us to identify eight hub genes constituting a CMLS with robust prognostic predictive power. Comprehensive analysis of multi-omics data can provide critical insights and further refine the molecular classification of LUSC. The identification of MoSs represents a valuable tool for early prognostic prediction of patients and for screening potential candidates likely to benefit from immunotherapy, with broad implications for clinical practice. In conclusion, our study offers novel insights into LUSC molecular subtypes based on comprehensive multi-omics data and encourages precise treatment of LUSC patients in the future.</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

Integrated multiomics analysis and advanced machine learning techniques to refine molecular subtypes, stratify prognosis, characterize tumor microenvironment, and identify distinct sensitivity patterns to frontline therapies in lung squamous cell carcinoma

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月24日

查看全文
  • Shi Yan,
  • Xu Zhang,
  • Tongying Xu,
  • Xiangyu Wang,
  • Fanglin Tian,
  • Weina Fan,
  • Yuning Zhan,
  • Li Cai,
  • Ying Xing

摘要

Lung squamous cell carcinoma (LUSC) is characterized by high aggressiveness and significant heterogeneity, currently lacking highly precise individualized treatment options. Leveraging computational methodologies, we integrated multi-omics data from LUSC patients using 10 clustering algorithms, coupled with 10 machine learning algorithms, to delineate high-resolution molecular subtypes and develop a robust consensus machine learning-driven signature (CMLS). Through multi-omics clustering, we identified two multiomics subtypes (MoSs) associated with prognosis, with MoS1 exhibiting the most favorable prognostic outcome. MoS2 demonstrated a heightened immune-infiltrated state and a greater likelihood of responding to immunotherapy, while MoS1, although potentially less responsive to immunotherapy, may benefit more from Erlotinib. Moreover, we sought potential subchemotherapeutic targets based on the molecular expression characteristics of MoSs. Subsequent screening enabled us to identify eight hub genes constituting a CMLS with robust prognostic predictive power. Comprehensive analysis of multi-omics data can provide critical insights and further refine the molecular classification of LUSC. The identification of MoSs represents a valuable tool for early prognostic prediction of patients and for screening potential candidates likely to benefit from immunotherapy, with broad implications for clinical practice. In conclusion, our study offers novel insights into LUSC molecular subtypes based on comprehensive multi-omics data and encourages precise treatment of LUSC patients in the future.