基因表現調控演算是個高度挑戰性的研究問題,同側因子(Cis-elements)調控序列演算分析逐漸展現建構基因組整體表現調控網絡的重要性。本論文提出演算選設與組合選群方法,建構序列上游調控區同側元件組合的基因調控組樹。首先探索產生對側因子(Trans-factors)直接結合核苷酸序列的常規表示(regular expression)模式演算選設同側因子(Cis-elements)樣本。憑藉生物實驗確認的同側調控序列 (Motifs)完整集合群公知序列資料庫EPD[15]及TRANSFAC[21]挑選資料,作為最佳化收斂同側因子調控序列的自然演化選用機制本質依歸。 基因調控組樹建構採用演算選設同側因子調控元件群執行標註索引序列上游調控區方法,階層式資料分群演算輸出基因調控組的系統樹圖型。基因調控組樹(RegTree)反映基因間調控模式的相似程度,是穩定性基因轉錄表現相對關係,可能是動態性演算模擬調控網絡(RegNet)的可行參考基礎。基因調控組樹可以協尋基因微陣相近表現型式基因歸群的對側調控因子,依據調控組樹歸群的群內與群間同側因子的異同,蒐尋可能的同側因子進行生物實驗確認對側轉錄因子。穩定性基因調控組樹或能未來應用協助演算模擬動態性基因轉錄表現調控網絡的完整過程。
The regulation of gene expression is a challenging problem. In the post-genomic era, the analysis of cis-regulation is growing importance. We propose a construct methodology as of the regulomics tree (RegTree) by using a putative set of simplistic cis-elements and algorithms with regulation expression. The cis-elements are created by mining the gapped patterns of direct contacting nucleotides within a selected data set from the public database, EPD[15] and TRANSFAC[21]. The putative set of simplistic cis-elements has been used to construct the regulomics tree by indexing cis-elements in the upper stream control region of gene sequences and data clustering method. By using the hierarchical clustering algorithm, the output is a dendrogram that is desired in the regulomics tree. The regulomics tree reflects the similarity of regulation patterns among genes and is the reference model of dynamic simulation for RegNet. It can do help for microarray experiment to find the transcription factors. The application of regulomics tree is able to deduce that the overall program of the evolution from fertilized egg to individual body in the future.