Limits...
Identifying module biomarkers from gastric cancer by differential correlation network

View Article: PubMed Central - PubMed

ABSTRACT

Gastric cancer (stomach cancer) is a severe disease caused by dysregulation of many functionally correlated genes or pathways instead of the mutation of individual genes. Systematic identification of gastric cancer biomarkers can provide insights into the mechanisms underlying this deadly disease and help in the development of new drugs. In this paper, we present a novel network-based approach to predict module biomarkers of gastric cancer that can effectively distinguish the disease from normal samples. Specifically, by assuming that gastric cancer has mainly resulted from dysfunction of biomolecular networks rather than individual genes in an organism, the genes in the module biomarkers are potentially related to gastric cancer. Finally, we identified a module biomarker with 27 genes, and by comparing the module biomarker with known gastric cancer biomarkers, we found that our module biomarker exhibited a greater ability to diagnose the samples with gastric cancer.

No MeSH data available.


ROC curves obtained with our module markers and published biomarkers.Notes: (A) The gene expression dataset used to identify differential network, and three independent datasets GSE13911 (B), GSE2701 (C), and GSE19826 (D) that were not used in this work.Abbreviations: AUC, area under the curve; ROC, receiver operating characteristic.
© Copyright Policy
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5036598&req=5

f5-ott-9-5701: ROC curves obtained with our module markers and published biomarkers.Notes: (A) The gene expression dataset used to identify differential network, and three independent datasets GSE13911 (B), GSE2701 (C), and GSE19826 (D) that were not used in this work.Abbreviations: AUC, area under the curve; ROC, receiver operating characteristic.

Mentions: In order to further validate the predicted gastric cancer genes, we applied these genes to separate cancer samples from normal samples. Three independent gene expression (microarray) datasets (GSE19826, GSE2701, and GSE13911) of gastric cancer were obtained from GEO database (http://www.ncbi.nlm.nih.gov/geo/) to validate the ability of classification of the module biomarker. At the same time, an RNA-Seq dataset of gastric cancer with 33 normal samples and 183 tumor samples was downloaded from TCGA database (http://cancergenome.nih.gov) to validate the ability of classification for the module biomarker on the differential data type. If these genes can successfully separate cancer samples from control samples for gastric cancer, we predicted that these genes can be proven to be related to gastric cancer. We used the five-fold cross-validation for SVM to detect the ability of classification for our predicted module biomarkers, and the AUC of ROC curve can be used to evaluate the ability of classification for biomarkers, so we can see that the AUCs of ROC curves for three of the four gastric cancer datasets are >0.9 and only one AUC is 0.88 for one dataset (Figure 5), and the AUC of ROC curves for the RNA-Seq data was also >0.9 (Figure 6). It means that our module biomarkers have a strong ability to identify gastric cancer and that these biomarker genes are strongly related to gastric cancer.


Identifying module biomarkers from gastric cancer by differential correlation network
ROC curves obtained with our module markers and published biomarkers.Notes: (A) The gene expression dataset used to identify differential network, and three independent datasets GSE13911 (B), GSE2701 (C), and GSE19826 (D) that were not used in this work.Abbreviations: AUC, area under the curve; ROC, receiver operating characteristic.
© Copyright Policy
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5036598&req=5

f5-ott-9-5701: ROC curves obtained with our module markers and published biomarkers.Notes: (A) The gene expression dataset used to identify differential network, and three independent datasets GSE13911 (B), GSE2701 (C), and GSE19826 (D) that were not used in this work.Abbreviations: AUC, area under the curve; ROC, receiver operating characteristic.
Mentions: In order to further validate the predicted gastric cancer genes, we applied these genes to separate cancer samples from normal samples. Three independent gene expression (microarray) datasets (GSE19826, GSE2701, and GSE13911) of gastric cancer were obtained from GEO database (http://www.ncbi.nlm.nih.gov/geo/) to validate the ability of classification of the module biomarker. At the same time, an RNA-Seq dataset of gastric cancer with 33 normal samples and 183 tumor samples was downloaded from TCGA database (http://cancergenome.nih.gov) to validate the ability of classification for the module biomarker on the differential data type. If these genes can successfully separate cancer samples from control samples for gastric cancer, we predicted that these genes can be proven to be related to gastric cancer. We used the five-fold cross-validation for SVM to detect the ability of classification for our predicted module biomarkers, and the AUC of ROC curve can be used to evaluate the ability of classification for biomarkers, so we can see that the AUCs of ROC curves for three of the four gastric cancer datasets are >0.9 and only one AUC is 0.88 for one dataset (Figure 5), and the AUC of ROC curves for the RNA-Seq data was also >0.9 (Figure 6). It means that our module biomarkers have a strong ability to identify gastric cancer and that these biomarker genes are strongly related to gastric cancer.

View Article: PubMed Central - PubMed

ABSTRACT

Gastric cancer (stomach cancer) is a severe disease caused by dysregulation of many functionally correlated genes or pathways instead of the mutation of individual genes. Systematic identification of gastric cancer biomarkers can provide insights into the mechanisms underlying this deadly disease and help in the development of new drugs. In this paper, we present a novel network-based approach to predict module biomarkers of gastric cancer that can effectively distinguish the disease from normal samples. Specifically, by assuming that gastric cancer has mainly resulted from dysfunction of biomolecular networks rather than individual genes in an organism, the genes in the module biomarkers are potentially related to gastric cancer. Finally, we identified a module biomarker with 27 genes, and by comparing the module biomarker with known gastric cancer biomarkers, we found that our module biomarker exhibited a greater ability to diagnose the samples with gastric cancer.

No MeSH data available.