利用彈性網路分子動力學研究蛋白質動態

Abstract

在結構計算生物學上，蛋白質的動態一直與其功能有密切的關係。而在實驗上ｘ光結晶繞射所得的溫度因數與核磁共振所得到的蛋白質動態多重模型也告知我們蛋白質並非靜態的。在過去，我們利用分子動力學來研究蛋白質動態，但是分子動力學的時間複雜度太過龐大且須巨大的計算量。在這篇論文中，首先我們試著建立一個粗略的模型來簡化分子動力學中過多的力學因子與及原子個數來計算分子的動態軌跡然後我們將其的振動來跟ｘ光結晶繞射所得的溫度因數作比較。再來我們試著考慮各參數對這個模型的影響。並針對一個資料集合做整體的測試並分析其少數差距較大的蛋白質。最後我們證實蛋白質骨架的振動其實與其胺基酸種類並無太大的關係，而是由結構體決定其振動大小。

In computational structural biology, protein motion has relationships about function. Some experimental evidence such as temperature factor (B factor) or NMR multi-structure also shows that protein is not static. In traditionally, the molecular dynamics is useful to studying the protein motion, but there are so huge time complexity and computational scale on MD. In this article, first, we try to generate a coarse-grained model to simplify the force characteristics and simulate the trajectory of protein dynamics. Then we compare the result of fluctuation with B-factor from X-ray crystallography protein structure. Then we try to change the parameter to test the model and calculate a whole dataset. At last, we found that the protein fluctuation is decided by the structure.