Harbin Medical University, China.
Comput Biol Med. 2010 Feb;40(2):159-67. doi: 10.1016/j.compbiomed.2009.11.012. Epub 2009 Dec 31.
The aberrant hypermethylation of CpG islands in promoter regions of genes plays an important role in the onset and progression of Breast cancer. Meanwhile, it is highly associated with human genomic features. Two feature selection algorithms: t-test and CfsSubsetEval were used to obtain efficient feature subsets. We discovered 14 significant feature subsets by CfsSubsetEval, which can distinguish hypermethylated genes from control genes. As a result, 393 unconfirmed hypermethylated genes in Breast cancer were prioritized. These genes were assigned the hypermethylated scores and were supported by literature and Gene Ontology enrichment. This paper suggests that the feature subsets could be served as discriminating genomic markers to infer novel hypermethylated genes in cancer potentially.
CpG 岛启动子区域中基因的异常高甲基化在乳腺癌的发生和发展中起着重要作用。同时，它与人的基因组特征高度相关。我们使用了两种特征选择算法：t 检验和 CfsSubsetEval，以获得有效的特征子集。通过 CfsSubsetEval，我们发现了 14 个显著的特征子集，可以区分高甲基化基因和对照基因。结果，我们优先确定了 393 个乳腺癌中未经证实的高甲基化基因。这些基因被赋予了高甲基化评分，并得到了文献和基因本体富集的支持。本文表明，这些特征子集可以作为区分基因组标记，潜在地推断癌症中的新的高甲基化基因。
