Selection of gastric cancer subgroups marker genes based on machine learning methods - 文章详情页

作者：

Li, J.-G. (Li, J.-G..) | Li, P. (Li, P..) | Yan, Z. (Yan, Z..) | Li, J. (Li, J..) | Ruan, X.-G. (Ruan, X.-G..)

收录：

Scopus PKU CSCD

摘要：

Using　machines　learning　methods　to　find　new　gastric　cancer　biomarkers　provides　us　a　standard　and　basis　for　exploring　the　molecular　mechanisms　of　gastric　cancer　and　the　diagnosis　and　cure　of　gastric　cancer　from　gene　level.　We　employed　33　Oligo　gene　chips　microarray　dataset　of　Chinese,　including　13　diffused　gastric　samples　and　20　intestinal　gastric　samples.　And　each　of　the　samples　had　21378　genes.　A　hybrid　method,　including　the　significant　analysis　of　microarrays　(SAM),　the　partial　least　squares　(PLS)　and　Bhattacharyya　distance-sequence-forward　search　(BD-SFS),　was　used　to　reduce　the　dimensions　of　the　data.　20　genes　were　selected　as　feature　genes　at　last.　The　SVM　classifier　could　distinguish　diffused　ones　and　intestinal　ones　well　by　using　these　20　genes　data.　The　accuracy　rate　reached　89.45%.　And　the　classification　accuracy　rate　of　hierarchical　clustering　could　reach　at　93.94%.　In　addition,　biological　significance　analysis　showed　that　most　of　these　20　genes　were　important　for　the　diagnosis　and　molecular　classification　of　some　human　malignant　tumors.

关键词：

Feature selection; Gastric cancer; Gene expression profile; Marker genes

作者机构：

[ 1 ] [Li, J.-G.]Academy of Electronic Information and Control Engineering, Beijing University of Technology, Beijing 100124, China
[ 2 ] [Li, P.]Academy of Electronic Information and Control Engineering, Beijing University of Technology, Beijing 100124, China
[ 3 ] [Yan, Z.]Pehing University School of Oncology, Beijing Cancer Hospital and Institute, Beijing 100142, China
[ 4 ] [Li, J.]Basic Ddepartment, Wuhan Military Economic College in Xiangfan, Xiangfan 441118, China
[ 5 ] [Ruan, X.-G.]Academy of Electronic Information and Control Engineering, Beijing University of Technology, Beijing 100124, China

通讯作者信息：

[Li, J.-G.]Academy of Electronic Information and Control Engineering, Beijing University of Technology, Beijing 100124, China

电子邮件地址：

lijg@bjut.edu.cn

查看成果更多字段

成果类型
所属机构

所有年份指定年份从至