| At line 1 changed 2 lines |
| __MarkRank__: Prioritization of network biomarkers for complex diseases |
| * Last updated: August 8, 2016 |
| __MarkRank__: Discovering cooperative biomarkers for heterogeneous complex disease diagnoses |
| * Last updated: January 16, 2017 |
| At line 6 changed one line |
| Biomarkers with high reproducibility and accurate prediction performance can contribute to comprehending the underlying pathogenesis of related complex diseases, which can further facilitate disease diagnosis and therapy. With the powerful technologies stemming from systems biology, the types of disease-related biomarkers have been gradually refocused from conventional individual genes to current network-based disease biomarker modules. Techniques integrating both expression profiles and biological networks for the identification of connected disease biomarkers are receiving increasing interest. However, the incompleteness of protein-protein interaction (PPI) networks and the intrinsic heterogeneity of complex diseases may hinder the efficacy of connected biomarker identification. In this paper, we propose a novel method, MarkRank, to prioritize disease genes by integrating multi-source information including human PPI network, prior information about related diseases, and the discriminative power of cooperative gene combinations. In comparison with well-designed simulation studies and real biological datasets, MarkRank, explicitly taking the heterogeneity of diseases into consideration, achieves a superior performance than existing methods, exhibits high specificity associated with the related diseases, and can help exploring the underlying pathogenesis of complex disease in a new perspective. MarkRank has been implemented in the R package Corbi, which can be readily installed and used in R. |
| Biomarkers with high reproducibility and accurate prediction performance can contribute to comprehending the underlying pathogenesis of related complex diseases and further facilitate disease diagnosis and therapy. Techniques integrating gene expression profiles and biological networks for the identification of network-based disease biomarkers are receiving increasing interest. In this study, we proposed a novel method, MarkRank, to discover cooperative biomarkers for heterogeneous diseases by integrating multi-source information, including biological networks, prior information about related diseases, and the cooperative effects of gene combinations. By innovatively constructing a gene cooperation network to capture the cooperative effects of gene combinations and explicitly taking the heterogeneity of complex diseases into consideration, MarkRank achieves superior performance compared to existing methods in both simulation studies and real datasets. The biomarkers identified by MarkRank not only have a better prediction accuracy in the Monte Carlo cross-validation procedure but also have stronger topological relationships in the biological network and exhibit high specificity associated with the related diseases. Furthermore, the top genes identified by MarkRank involve crucial biological processes of related diseases and give a good prioritization for known disease genes. In conclusion, the MarkRank method is an efficient and effective computational tool for discovering disease biomarkers and can help in the exploration of the underlying pathogenesis of complex diseases. MarkRank has been implemented in the R package Corbi, which is publicly available at the CRAN website (http://cran.r-project.org/web/packages/Corbi/). The open source codes, original datasets and additional materials are available at http://doc.aporc.org/wiki/MarkRank. |
| At line 10 changed one line |
| * Duanchen Sun, Xianwen Ren, Eszter Ari, Tamas Korcsmaros, Peter Csermely, Ling-Yun Wu. MarkRank: Prioritization of network biomarkers for complex diseases. In submission, 2016. |
| * Duanchen Sun, Xianwen Ren, Eszter Ari, Tamas Korcsmaros, Peter Csermely, Ling-Yun Wu. Discovering cooperative biomarkers for heterogeneous complex disease diagnoses. In submission, 2017. |
