搜尋蛋白質中相關聯殘基所構成的網絡

Abstract

同源蛋白的多重序列排比帶有演化紀錄，從這些紀錄中可以分析出許多有用的資訊。這類分析中最常見的即是保守性分析；一個排比的位置如果總是只能觀察到一種胺基酸殘基，我們就稱它是保守性十分強的位置。保守性強的位置通常是蛋白質中行使功能的區域，這是因為功能的行使會限制住該區域可能產生的突變。 然而，有些位置雖然保守性不強，卻仍然是蛋白質行使功能的重要區域。胺基酸之間的關聯性解釋了這種現象。許多研究者表示，要了解蛋白質之中胺基酸間的關聯性，可從對兩個位置作關聯突變分析得知。 在這份研究中，我們利用保守性熵和關聯性熵來預測相關聯的殘基並進行比較。我們的研究初步證實經校正的關聯性熵是有效的預測關聯突變的指數。我們同時也提出一種新的空間搜尋方法來找出關聯殘基所組成的網絡和它的意義。

Multiple sequence alignments (MSA) carry evolutionary records. By analyzing these records, one may get useful biological information. A common analysis method is conservation analysis. If there is no record in mutation at one MSA position, we say that position tends to be conserved. Conserved positions are always functional important sites since functionality imposes constraints on these sites from mutation. Although some positions show little conservation, they are still functionally important. The coupling residues account this observation. Many investigators have proposed that analyzing correlated mutation in two positions can effectively predict the coupling degree between these two residues. Thus, in this study, we introduce conservation entropy and coupling entropy to detect coupling between two sites and also compare the effect between two. Our study generally confirmed the effect of normalized coupling entropy to be a good indicator in detecting correlated mutation. We also introduce a network searching strategy to verify the identity of coupling residues network