A novel chi-square statistic for detecting group differences between pathways in systems epidemiology.
Traditional epidemiology often pays more attention to the identification of a single factor rather than to the pathway that is related to a disease, and therefore, it is difficult to explore the disease mechanism. Systems epidemiology aims to integrate putative lifestyle exposures and biomarkers ext...
| Publicado en: | Statistics in Medicine Vol. 35; no. 29; pp. 5512 - 5525 |
|---|---|
| Autores principales: | , , , , , , |
| Formato: | equations & formulas research tables/charts Journal Article |
| Publicado: |
Wiley-Blackwell
12/20/2016
|
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=119354279&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 119354279 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 02776715 2DZ jtl: Statistics in Medicine issn: 02776715 maglogo: Y pubinfo: dt: 12/20/2016 vid: 35 iid: 29 pid: 480 pub: Wiley-Blackwell place: Malden, Massachusetts artinfo: ui: 119354279 119354279 NLM27605026 119354279 10.1002/sim.7094 NLM27605026 119354279 ppf: 5512 ppct: 13 formats: tig: atl: A novel chi-square statistic for detecting group differences between pathways in systems epidemiology. aug: au: Yuan, Zhongshang Ji, Jiadong Zhang, Tao Liu, Yi Zhang, Xiaoshuai Chen, Wei Xue, Fuzhong affil: Department of Biostatistics, School of Public Health, Shandong University, Jinan 250012, Shandong, China sug: subj: Chi Square Test Prospective Studies Epidemiological Research Human ab: Traditional epidemiology often pays more attention to the identification of a single factor rather than to the pathway that is related to a disease, and therefore, it is difficult to explore the disease mechanism. Systems epidemiology aims to integrate putative lifestyle exposures and biomarkers extracted from multiple omics platforms to offer new insights into the pathway mechanisms that underlie disease at the human population level. One key but inadequately addressed question is how to develop powerful statistics to identify whether one candidate pathway is associated with a disease. Bearing in mind that a pathway difference can result from not only changes in the nodes but also changes in the edges, we propose a novel statistic for detecting group differences between pathways, which in principle, captures the nodes changes and edge changes, as well as simultaneously accounting for the pathway structure simultaneously. The proposed test has been proven to follow the chi-square distribution, and various simulations have shown it has better performance than other existing methods. Integrating genome-wide DNA methylation data, we analyzed one real data set from the Bogalusa cohort study and significantly identified a potential pathway, Smoking → SOCS3 → PIK3R1, which was strongly associated with abdominal obesity. The proposed test was powerful and efficient at identifying pathway differences between two groups, and it can be extended to other disciplines that involve statistical comparisons between pathways. The source code in R is available on our website. Copyright © 2016 John Wiley & Sons, Ltd. pubtype: Academic Journal doctype: equations & formulas research tables/charts Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|