Computational Approaches for Prioritizing Disease Candidate Genes and Discovering Pathway Cross-talks

• Main Authors: Chia-Lang Hsu, 許家郎
• Other Authors: Ueng-Cheng Yang
• Format: Others
• Language: en_US
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/58526207336106522150

博士 === 國立陽明大學 === 生物醫學資訊研究所 === 100 === Because of the rapid accumulation of genomic data and the complexity of biological systems, computational support in experiment design, processing of results, and interpretation of results has become essential in current biomedical research. In this dissertation, we proposed computational approaches to resolve the problems that have been encountered in biomedical research. Associating genes with diseases is required to understand disease mechanisms, diagnosis, and therapy. At present, methods such as linkage analysis and genome-wide association study can identify the regions in which unknown disease genes are located, but the regions could contain up to hundreds of candidate genes. On the basis of the assumption that genes related to similar disease phenotype tend to be located in neighborhood in biological networks, numerous network-based methods have developed for prioritizing the candidate genes. In this thesis, interconnectedness (ICN) score was proposed for ranking the candidate genes. This ICN-based method is not only comparable with well-known methods, but also outperforms other methods when a candidate disease gene is not directly linked to known disease genes in a network. Interestingly, we show that a combined scoring strategy could enable ICN to achieve an even better performance than other methods used alone. In biological systems, pathways coordinate or interact with one another to achieve complex biological processes. Studying how they influence each other is essential for understanding the intricacies of a biological system. However, current methods rely on statistical tests to determine pathway relations, and may lose numerous biologically significant relations. To address this problem, a function-based approach (FBA) was proposed to identify the pathway cross-talks. This method uses the Gene Ontology annotations of pathway components to measure the functional similarities between pathways. The performance of this method was better than the existing methods based on the shared components and protein-protein interactions. Many well-known pathway cross-talks are only identified by FBA. In addition, the false positive rate of FBA is significantly lower than the others via pathway co-expression analysis. This proposed method appears to be more sensitive and able to infer more biologically significant and explainable pathway relations.