- 學位論文
以RNA二級結構搜尋內轉譯子之研究
Searching for Internal Ribosome Entry Site in RNA secondary structure
摘要
IRES擁有核醣體內部進入的特點，可以啟動與傳統不同的轉譯起始途徑。這個特性除了提供控制轉譯的方法，同時也是RNA不轉譯區調控基因表現的一個例子。讓生物學家感興趣的是，究竟還有多少在結構上類似已知IRES的序列存在序列資料庫之中？本研究使用生物資訊解析RNA結構，希望幫助生物學家找出這個問題的答案。 本研究探討目前預測及比較RNA二級結構的方法，並且串接現有的RNA生物資訊工具，完成搜尋RNA二級結構的系統。系統首先利用RNALfold程式，由一級序列預測小區域內較穩定的RNA二級結構，再使用RNA Align程式，將預測的結構比對已知的IRES結構，藉此搜尋潛在的IRES序列。 本研究為瞭解系統的特性及評估系統搜尋IRES的能力。首先測試在四個已知具有IRES的病毒全基因中尋找長度為205個鹼基的腸病毒71型IRES Domain IV，被搜尋的目標包含腸病毒71型、牛腸病毒、鼻病毒以及C型肝炎病毒，結果顯示這樣的設計能夠找到腸病毒屬IRES序列的位置。 另外針對UTRdb Virus 5’ UTR搜尋長度為206個鹼基的C型肝炎病毒IRES Domain III，結構預測程式的最大預測結構長度參數分別設定為100，250以及400個鹼基，隨後進行結果的正確率統計，結果發現當最大結構長度參數設為250時，能夠找到最多的HCV以及Pestivirus IRES。此外，在去除黃病毒科的結果以後，發現Simian Picornavirus 12以及Porcine Enterovirus 8在5’端UTR有結構與搜尋目標相似，而且序列比對的相似度可達40%的結構出現，這些序列片段是否真的具有與HCV IRES相同的功能，可待更進一步的實驗證實。
並列摘要
IRES (Internal Ribosome Entry Site) provides an alternative, cap-independent translation initiation pathway. This is a typical case of RNA untranslated region (UTR) regulates gene expression. It is interesting to the biologist: How many sequences in the database are structurally similar to those known IRES? In this research, a series of RNA bioinformatics tools have been applied to answer this question. After reviewing the contemporary RNA secondary structure prediction and comparison methods, two RAN bioinformatics tools were implemented to perform two-stage strategy IRES searching flow. In the first stage, the RNALfold program is used to predict locally stable RNA secondary structures. The RNA Align was followed to compare the predefined IRES to the aforementioned RNA structures. The potential IRES regions were screened by the cascaded workflow. To evaluate the workflow performance, four virus complete genomes, including EV71 USA MS strain, Bovine Enterovirus, Human Rhinovirus and Hepatitis C Virus (HCV) were used to searching for IRES domain IV of Enterovirus type 71 (EV71, contain 205 nucleotides). The results shown the workflow can successfully find the IRES of Enterovirus genus. Searching the HCV IRES domain III, which contain 206 nucleotides, in UTRdb virus 5' UTR sequences was also conducted. The prediction sensitivity with the setting of the maximum allowable predicted structure length to 100, 250 and 400 nucleotides was calculated. The results shown that when the length was set to 250, one can find most HCV and Pestivirus IRESes. In addition, it is also found that there are possible target IRES structure in the 5' UTR of Simian Picornavirus 12 (SV45) and Porcine Enterovirus 8 (PEV-8). However, the IRES activity and functional similarity on these regions require further proven by wet-lab experiments.