利用代理人模型模擬CHO細胞在靜態培養下的生長

Abstract

CHO細胞為中國倉鼠卵巢細胞，其具穩定生長特性，廣泛應用於重組蛋白與單株抗體的生產。CHO細胞培養在24孔盤中是常見的培養模式，製藥產業中亦常見以此種培養方式，作為放大培養之前的早期培養階段，用以產生所需的蛋白。 因為24孔盤的半封閉幾何形狀以及擴散限制，此種培養形式無法快速補充培養液中的氧氣濃度，故在細胞數目隨時間快速增長的情形下，細胞容易遭遇缺氧環境，進而影響代謝。 本論文利用Julia程式語言實作一代理人模型，以細胞週期為基礎，搭配不同環境條件下的代謝趨勢，模擬典型CHO細胞在24孔盤中的培養情形，並觀察缺氧情形下幾個重要的量測指標。

Chinese hamster ovary (CHO) cells are widely used in the production of recombinant proteins and monoclonal antibodies due to their stable growth characteristics. Culturing CHO cells in 24-well plates is a common practice, particularly in the pharmaceutical industry, where it serves as an early-stage system prior to large-scale bioreactor cultivation. However, the semi-enclosed geometry and diffusion limitations of 24-well plates restrict oxygen replenishment in the culture medium. As cell density increases, these limitations often lead to hypoxic conditions, which in turn affect cell metabolism and growth dynamics. In this thesis, the Julia programming language is used to develop an agent-based model that integrates cell cycle regulation with metabolic responses under varying environmental conditions. The model simulates CHO cell culture in 24-well plates and enables the analysis of key indicators of cell behavior under hypoxia, providing insights that may guide experimental design and bioprocess optimization.