Scholarlink

<p>Phenotypically healthy cells frequently harbor somatic variants at cancer-associated genes, indicating that malignant transformation requires the selection of several alterations. Predicting which combinations of mutations, or co-mutations, exhibit oncogenic capacity requires identifying co-mutations that occur more or less frequently than expected. However, statistical frameworks to solve this problem are hampered by tumor heterogeneity and data availability. Here we curated putative oncogenic mutations in &gt;70,000 human tumors from 119 subtypes, and designed a strategy to search for co-mutations based on in silico simulation of mutagenesis (SelectSim). Using this dataset and tool, we discovered and validated co-mutations across independent human cohorts, compared co-mutations across different tumor types and identified potential risk factors of metastatic progression. Notably, across several cohorts of phenotypically normal tissue samples, we show that, unlike individual oncogenic variants, significantly co-occurring mutations are largely cancer-specific and are observed rarely in healthy tissues, providing clues about the paths to tumorigenesis.</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

Evolving patterns of co-mutations from tumor initiation to metastatic progression

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月10日

查看全文
  • Arvind Iyer,
  • Miljan Petrovic,
  • Debora Sesia,
  • Luca Nanni,
  • Marco Mina,
  • Giovanni Ciriello

摘要

Phenotypically healthy cells frequently harbor somatic variants at cancer-associated genes, indicating that malignant transformation requires the selection of several alterations. Predicting which combinations of mutations, or co-mutations, exhibit oncogenic capacity requires identifying co-mutations that occur more or less frequently than expected. However, statistical frameworks to solve this problem are hampered by tumor heterogeneity and data availability. Here we curated putative oncogenic mutations in >70,000 human tumors from 119 subtypes, and designed a strategy to search for co-mutations based on in silico simulation of mutagenesis (SelectSim). Using this dataset and tool, we discovered and validated co-mutations across independent human cohorts, compared co-mutations across different tumor types and identified potential risk factors of metastatic progression. Notably, across several cohorts of phenotypically normal tissue samples, we show that, unlike individual oncogenic variants, significantly co-occurring mutations are largely cancer-specific and are observed rarely in healthy tissues, providing clues about the paths to tumorigenesis.