Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD 4072, Australia.
Bioinformatics. 2013 Jun 15;29(12):1553-61. doi: 10.1093/bioinformatics/btt191. Epub 2013 Apr 23.
Deciphering the modus operandi of dysregulated cellular mechanisms in cancer is critical to implicate novel cancer genes and develop effective anti-cancer therapies. Fundamental to this is meticulous tracking of the behavior of core modules, including complexes and pathways across specific conditions in cancer.
Here, we performed a straightforward yet systematic identification and comparison of modules across pancreatic normal and cancer tissue conditions by integrating PPI, gene-expression and mutation data. Our analysis revealed interesting change-patterns in gene composition and expression correlation particularly affecting modules responsible for genome stability. Although in most cases these changes indicated impairment of essential functions (e.g., of DNA damage repair), in several other cases we noticed strengthening of modules possibly abetting cancer. Some of these compensatory modules showed switches in transcription regulation and recruitment of tumor inducers (e.g., SOX2 through overexpression). In-depth analysis revealed novel genes in pancreatic cancer, which showed susceptibility to copy-number alterations (e.g., for USP15 in 17 of 67 cases), supported by literature evidence for their involvement in other tumors (e.g., USP15 in glioblastoma). Two of the identified genes, YWHAE and DISC1, further supported the nexus between neural genes and pancreatic carcinogenesis. Extension of this assessment to BRCA1 and BRCA2 breast tumors showed specific differences even across the two sub-types and revealed novel genes involved therein (e.g., TRIM5 and NCOA6).
Our software CONTOURv1 is available at: http://bioinformatics.org.au/tools-data/.
解析癌症中失调细胞机制的作用方式对于牵连新的癌症基因和开发有效的抗癌疗法至关重要。为此，至关重要的是仔细跟踪核心模块（包括复合物和途径）在癌症特定条件下的行为。
在这里，我们通过整合 PPI、基因表达和突变数据，对胰腺正常和癌症组织条件下的模块进行了直接而系统的识别和比较。我们的分析揭示了基因组成和表达相关性的有趣变化模式，特别是影响负责基因组稳定性的模块。尽管在大多数情况下，这些变化表明基本功能受损（例如，DNA 损伤修复），但在其他几种情况下，我们注意到模块的加强可能有助于癌症。其中一些补偿模块显示转录调控和肿瘤诱导物（例如 SOX2 通过过表达）的招募开关。深入分析揭示了胰腺癌中的新基因，这些基因显示出对拷贝数改变的敏感性（例如，在 67 例中的 17 例中 USP15），并得到了它们参与其他肿瘤（例如，USP15 在神经胶质瘤中的证据的支持）。鉴定出的两个基因，YWHAE 和 DISC1，进一步支持了神经基因与胰腺癌发生之间的联系。将这种评估扩展到 BRCA1 和 BRCA2 乳腺癌，即使在两种亚型之间也显示出特定的差异，并揭示了其中涉及的新基因（例如，TRIM5 和 NCOA6）。
我们的软件 CONTOURv1 可在以下网址获得：http://bioinformatics.org.au/tools-data/。
