| Info. | Vol.9 - No.4 (2015.12.20) |
|---|---|
| Title | NetRanker: A Network-based Gene Ranking Tool Using Protein-Protein Interaction and Gene Expression Data |
| Authors | Erkhembayar Jadamba1, Seong Beom Cho3 & Miyoung Shin2,* |
| Institutions | 1Bio-Intelligence & Data Mining Laboratory, Graduate School of Electrical Engineering and Computer Science, Kyungpook National University, 1370 Sankyuk-dong, Buk-gu, Daegu 702-701, Korea 2School of Electronics Engineering, Kyungpook National University, 1370 Sankyuk-dong, Buk-gu, Daegu 702-701, Korea 3Division of Bio-Medical Informatics, Center for Genome Science, Korea National Institute of Health, Korea |
| Abstract | Over the past years, many gene prioritization methods have been studied which mainly uses gene expression profiles, occasionally with other information like functional annotations. More recently, various biological network-based approaches are being widely explored to discover disease-related genes, showing the outperformance to earlier approaches via the guilt-by-association principle over the network topology. In this work, we developed a web-based tool of gene prioritization, called NetRanker, which allows users to identify the most significant top-n genes differentiating between two groups (e.g., case and control samples) by considering PPI networks as well as gene expression data. Particularly, this tool facilitates users to employ one of three different network-based ranking algorithms (including NR PageRank, NR-HeatDiff, and NR-HITs algorithms) in order to prioritize genes over the PPI network built from STRING or PINA databases. Here gene expression data are used to initialize the weights to the nodes of the PPI network, relying on the significance of differential gene expression. Based on these initial weights, their neighborhoods over a network are re-evaluated for gene prioritization. The input to NetRanker is allowed to have raw CEL files of gene expression data with the choice of PPI data source for network generation. Also, the initialization method can be chosen to determine initial weights of nodes over the network. To show the usefulness of this tool, we performed two case studies and one comparative study to identify disease-related genes. From our experimental results, it is observed that overall NR-PageRank performs better than NR-HeatDiff and NR-HITs, when algorithm parameters are chosen to equally reflect network connectivity and initial knowledge over network propagation. Also the network based ranking algorithms showed much better performance than earlier approaches. NetRanker is currently available to use in our webpage of http://biodmtool.knu.ac.kr/netranker. |
| Keyword | NetRanker, Gene prioritization, Networkbased ranking algorithm, Disease-related genes, Protein-protein interaction network |
| PDF File |
 # 2010년도 발행분 부터는 Springer 의 BioChip Journal 페이지에서 전문을 열람하실 수 있습니다.
# 학회회원 로그인 후 [ Springer BioChip Journal 열람하기 ] 버튼을 클릭하시면 새창으로 열립니다. |
