TY - JOUR
T1 - Identification of dilated cardiomyopathy signature genes through gene expression and network data integration
AU - Camargo, Anyela
AU - Azuaje, Francisco
N1 - Funding Information:
This work was supported in part by a grant from the EU-FP6, CARDIOWORKBENCH project, to F.A.
PY - 2008/12
Y1 - 2008/12
N2 - Dilated cardiomyopathy (DCM) is a leading cause of heart failure (HF) and cardiac transplantations in Western countries. Single-source gene expression analysis studies have identified potential disease biomarkers and drug targets. However, because of the diversity of experimental settings and relative lack of data, concerns have been raised about the robustness and reproducibility of the predictions. This study presents the identification of robust and reproducible DCM signature genes based on the integration of several independent data sets and functional network information. Gene expression profiles from three public data sets containing DCM and non-DCM samples were integrated and analyzed, which allowed the implementation of clinical diagnostic models. Differentially expressed genes were evaluated in the context of a global protein-protein interaction network, constructed as part of this study. Potential associations with HF were identified by searching the scientific literature. From these analyses, classification models were built and their effectiveness in differentiating between DCM and non-DCM samples was estimated. The main outcome was a set of integrated, potentially novel DCM signature genes, which may be used as reliable disease biomarkers. An empirical demonstration of the power of the integrative classification models against single-source models is also given.
AB - Dilated cardiomyopathy (DCM) is a leading cause of heart failure (HF) and cardiac transplantations in Western countries. Single-source gene expression analysis studies have identified potential disease biomarkers and drug targets. However, because of the diversity of experimental settings and relative lack of data, concerns have been raised about the robustness and reproducibility of the predictions. This study presents the identification of robust and reproducible DCM signature genes based on the integration of several independent data sets and functional network information. Gene expression profiles from three public data sets containing DCM and non-DCM samples were integrated and analyzed, which allowed the implementation of clinical diagnostic models. Differentially expressed genes were evaluated in the context of a global protein-protein interaction network, constructed as part of this study. Potential associations with HF were identified by searching the scientific literature. From these analyses, classification models were built and their effectiveness in differentiating between DCM and non-DCM samples was estimated. The main outcome was a set of integrated, potentially novel DCM signature genes, which may be used as reliable disease biomarkers. An empirical demonstration of the power of the integrative classification models against single-source models is also given.
KW - Biodata mining and integration
KW - Diagnostic systems
KW - Dilated cardiomyopathy
KW - Gene expression data
KW - Heart failure
KW - Protein networks
UR - http://www.scopus.com/inward/record.url?scp=56149092205&partnerID=8YFLogxK
U2 - 10.1016/j.ygeno.2008.05.007
DO - 10.1016/j.ygeno.2008.05.007
M3 - Article
C2 - 18595652
AN - SCOPUS:56149092205
SN - 0888-7543
VL - 92
SP - 404
EP - 413
JO - Genomics
JF - Genomics
IS - 6
ER -