Open AccessMining gene link information for survival pathway hunting
Author(s) -
Jing GaoJian,
Zhang Zirui,
Wang HongQiang,
Zheng HongMei
Publication year - 2015
Publication title -
iet systems biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.367
H-Index - 50
eISSN - 1751-8857
pISSN - 1751-8849
DOI - 10.1049/iet-syb.2014.0048
Subject(s) - proportional hazards model , survival analysis , statistic , gene , hazard ratio , link (geometry) , computational biology , test statistic , biology , computer science , data mining , bioinformatics , statistics , statistical hypothesis testing , genetics , mathematics , confidence interval , computer network
This study proposes a gene link‐based method for survival time‐related pathway hunting. In this method, the authors incorporate gene link information to estimate how a pathway is associated with cancer patient's survival time. Specifically, a gene link‐based Cox proportional hazard model (Link‐Cox) is established, in which two linked genes are considered together to represent a link variable and the association of the link with survival time is assessed using Cox proportional hazard model. On the basis of the Link‐Cox model, the authors formulate a new statistic for measuring the association of a pathway with survival time of cancer patients, referred to as pathway survival score (PSS), by summarising survival significance over all the gene links in the pathway, and devise a permutation test to test the significance of an observed PSS. To evaluate the proposed method, the authors applied it to simulation data and two publicly available real‐world gene expression data sets. Extensive comparisons with previous methods show the effectiveness and efficiency of the proposed method for survival pathway hunting.