蛋白質複合體交互作用網路之拓樸分析

Topological Analysis of Protein Complex Interaction Networks

陳登鴻(Teng-Hung Chen)

指導教授 ： 黃建宏

共同指導教授 ： 吳家樂

國立虎尾科技大學/電機資訊學院/資訊工程研究所/碩士(2015年)

https://doi.org/10.6827/NFU.2015.00073

摘要

複雜的網路系統在電腦科學、社會科學、分子與群體生物學等許多不同的領域扮演非常重要的角色。對於生物與醫學的領域來說，生物網路的分析應用廣泛，包括識別藥物標靶、確定蛋白質或基因的功能、設計有效策略診治疾病等。蛋白質複合體通常會與其他複合體產生交互作用形成蛋白質複合體交互作用網路（PCIN）參與重要的細胞過程。在本論文中，我們應用隨機圖形理論的方法於四種物種（human、mouse、rat、和yeast）的PCINs，從廣域的網路結構到區域的拓樸性質分析。我們有初步證據證明這四種PCINs是小世界網路，以及rat是屬於assortative network。
此外，我們利用Kolmogorov-Smirnov統計檢定四個物種的八種拓撲參數(接近度中心、分支度中心、離心度中心、參與中間度中心、特徵向量中心、橋接中心、群集係數、中間度係數及區域平均連通度)的累積頻率分布，結果顯示這些拓撲參數可以作為有效的指標來區分PCINs。透過scalability test也進ㄧ步驗證了我們研究的有效性，此外，使用拓樸性質的分類，我們找到了一些生物網路上重要的蛋白質複合體，而且證明這種拓樸分析的方法可以提供有意義的生物解釋。

關鍵字

蛋白質複合體 ； 蛋白質複合體交互作用網路 ； 拓樸性質 ； 網路結構 ； enrichment 分析

並列摘要

The theory of complex networks plays an important role in a wide variety of disciplines, ranging from computer science, sociology, engineering and physics, to molecular and population biology. Within the fields of biology and medicine, potential applications of network analysis include for example drug target identification, determining a protein’s or gene’s function, designing effective strategies for treating various diseases or providing early diagnosis of disorders. Protein complexes usually interact with other complexes to form protein complex interaction network (PCIN) that take part in important cellular processes. In this thesis, the random graph theory approach was applied to the study of PCINs of four species (human, rat, mouse and yeast), from the global perspective of their network structures to the zoomed-in view of local topological features. We proved that all four PCINs are small world networks, and the rat PCIN is significantly positively correlated to assortative networks.
The accumulative frequency distributions of the eight local topological parameters (closeness centrality, degree centrality, eccentricity centrality, betweenness centrality, eigenvector centrality, bridging centrality, clustering coefficient, brokering coefficient and local average connectivity) are different as a result of the pairwise tests among the four species using Kolmogorov-Smirnov statistics. This suggested that local topological parameters could be served as useful indictors to specify PCINs. We demonstrated the underlying network structures of the four PCINs, and the effectiveness of the current study is further investigated by the scalability test. Furthermore, by using the topological properties classification, we identified some critical protein complexes for the four PCINs, and validated that the topological analysis approach could provide meaningful biological interpretations of the protein complex interaction systems.

並列關鍵字

protein complexes ； protein complex interaction network ； topological property ； network structure ； enrichment analysis

參考文獻

[37] Konganti, K., Wang, G., Yang, E., Cai, J.J..: 'SBEToolbox: A Matlab Toolbox for Biological Network Analysis', Evolutionary Bioinformatics Online, 2013, 9, pp. 355-362.

[2] Chen, X.J., Squarr, A.J., Stephan, R., et al.: 'Ena/VASP proteins cooperate with the WAVE complex to regulate the actin cytoskeleton', Developmental Cell, 2014, 30(5), pp. 569-584.

[3] Antosiewicz, A., Senkara, E., Cieśla, J.: 'Quartz crystal microbalance with dissipation and microscale thermophoresis as tools for investigation of protein complex formation between thymidylate synthesis cycle enzymes', Biosensors & Bioelectronics, 2014, 64C, pp. 36-42.

[4] Tripsianes, K., Friberg, A., Barrandon, C., et al.: 'A novel protein-protein interaction in the RES (REtention and Splicing) complex', The Journal of Biological Chemistry, 2014, 289.

[5] Schmidt, C., Robinson, C.V.: 'A comparative cross-linking strategy to probe conformational changes in protein complexes', Nature Protocols, 2014, 9(9), pp. 2224-2236.