運用化學平衡解與酵素動力學於代謝網路模擬之效果分析

本研究利用離散型事件驅動方式模擬生化代謝反應，資料模型採修正後的Petri Nets，將基質與產物視為Place，將酵素視為Transition，濃度視為Token。Transition為一計算物件，有關酵素催化的反應速率之計算方式或濃度轉換方法將寫於其中，藉此來動態計算轉置兩方位置內的濃度值。定量研究酵素的催化性，稱為酵素動力學，亦即計算酵素的反應速率。影響酵素反應速率的因素很多，舉凡反應物濃度、酵素濃度或抑制劑等皆會影響，更增加計算反應速率的複雜性。通常生化模擬軟體在計算反應速率時，會把時間切成很多小等份，再來計算反應完成時會進行多少等份的時間，並從中觀察每個時間點上的濃度變化，但這樣會碰到一個問題，時間要切多細才夠精準？此種計算方式會耗用電腦大量的資源且模擬速度較慢。在本研究中，提出一種化學平衡解計算方法，此方法不需計算酵素的反應速率，而是利用平衡常數直接去求得一個反應的平衡解，再透過所有相連反應的觸發與濃度振盪，當一組代謝網路中的所有反應都達到平衡時，即為steady state。此法的優點是可以利用較少的參數即可模擬，且CPU的模擬時間會大幅降低。本研究以Java物件導向程式語言實作，最後以厭氧糖解作用(anaerobic glycolysis)為例，代入相同的數據，分別以化學平衡解和Michaelis-Menten酵素動力學來求取最後各反應物達steady state時的濃度值，結果顯示兩者的數據是近似的，且兩方法的濃度-時間圖亦近似，此結果說明在某些需求下，使用者可以用化學平衡解來取代Michaelis-Menten酵素動力學。

關鍵字

離散型事件驅動；化學平衡解；酵素動力學；代謝網路； Petri Nets

並列摘要

Enzyme dynamics and chemical equilibrium are two major concepts for metabolic pathway modeling and simulation. This thesis implements two algorithms for these concepts and plugs them into a Petri Nets based simulation software. Then we apply these algorithms to a same case and compare the simulation results. The pros and cons are also discussed to conclude the situations for which algorithm performs better. Enzyme dynamics is a quantity method about catalysis of enzyme. There are many factors affect the catalysis of enzyme such as concentration of reactants, enzymes and inhibitors. Enzyme dynamics is a time-slice method and it has a problem that is how to determine the period of time-slice. The above-mentioned problems cause enzyme dynamics more complex. The thesis implements an algorithm of chemical equilibrium that it doesn’t compute the velocity of enzyme. It is an algorithm of binary cut and can use a equilibrium constant to complete compute directly. Then when all the adjacent reactions are fired lastingly, it will be steady state. The advantage of this algorithm is that it uses lesser parameters and reduces the simulation time of CPU. Our implementation is based on Java language. After applying chemical equilibrium and Michaelis-Menten dynamics to simulate anaerobic glycolysis with the same data, we find the two solutions are similar. This information revealed that users can adopt chemical equilibrium to replace the Michaelis-Menten dynamics on some demands.